#Eurojust20 | 📂 High-profile cases from the last 20 years.
In 2020, Eurojust set up a joint investigation team (JIT) to help dismantle #EncroChat, 📱 an encrypted phone network used by criminals.
Thanks to the JIT, police listened to criminals plotting crimes & much more ⤵️
#Eurojust20 | 📂 High-profile cases from the last 20 years.
'Lost Boy' case (2010) ➡️ Eurojust facilitates the dismantling of a worldwide #ChildAbuse network in:
🇮🇹🇧🇪🇷🇴🇩🇪🇫🇷🇳🇱🇨🇿🇬🇧🇳🇴🇺🇸
Here's how we helped 👇
The Italian Desk at #Eurojust & the High School of the Judiciary held a seminar in Bari to
📈 evaluate the intensive cooperation of the 🇮🇹 judicial authorities w/ other 🇪🇺 Member States via Eurojust
🎈 celebrate the #Eurojust20 anniversary
Read more 👉 https://europa.eu/!DHbb6N
The intensive cooperation of the Italian judicial authorities with other EU Member States via Eurojust was evaluated, yesterday and today, during a dedicated seminar in Bari. The seminar, which also celebrated Eurojust’s 20th anniversary, was organised by the Bari seat of the High School of Judiciary and the Italian National Desk at the Agency. It aimed to enhance future ways of judicial cross-border cooperation in an increasingly digital era, focusing on particular crime areas such as cybercrime and terrorism.
📘 Today, we close our #Eurojust20 essay series with an epilogue:
'A forward look through the rear-view mirror'
- Ladislav Hamran, President of Eurojust
Thank you to all readers & contributors!
All 16 essays of the series ⤵️
https://europa.eu/!ch948B
Thinking in 2021 of how best to mark Eurojust’s 20th anniversary, I remember there was no shortage of ideas among colleagues. One of the suggestions we immediately grew fond of was to bring together written contributions from policymakers, academics and a rich assortment of colleagues to lay out how the story – not just of Eurojust, but of judicial cooperation in the European Union in general – has unfolded so far. Compiling this anniversary book has been an enriching exercise, as it sheds light on our work from many different angles at once. But even more than the different perspectives as such, it is the quality – without exception – of the contributions that sets this work apart. I could therefore not be more grateful to all authors for the time and effort they agreed to invest. And I could not be more pleased that the result of this collective effort has found its way to the publisher, and to you. Ultimately, the story of Eurojust is the result of the need for closer cooperation that was felt among judicial practitioners as the previous century was drawing to a close. In those days, prosecutors were often unsure where to send their letters rogatory but felt safe in their suspicions that a reply would not come soon – if at all. A watershed moment arrived when, in October 1999, the European Council in Tampere (Finland) concluded that To reinforce the fight against serious organised crime, the European Council has agreed that a unit (Eurojust) should be set up (…). Eurojust should have the task of facilitating the proper coordination of national prosecuting authorities and of supporting criminal investigations in organised crime cases (…). As demonstrated by Gilles de Kerchove’s insightful contribution on the genesis of Eurojust earlier in this book, it was an idea that had been pushed previously. History, however, confirms that European cooperation develops in stages rather than following a straight line, and Eurojust seeing the light of day was no exception. It only amplifies the gratitude we feel towards all those colleagues – to Gilles himself, for instance, but in equal measure to Hans Nilsson and many others – for the foresight and perseverance they have shown. Looking at the year-on-year development of the caseload Eurojust has been entrusted with by Member States, I am proud to say their vision quickly became reality. In its very first year and in a European Union of 15 Member States, Eurojust covered 202 files. The following years saw a steep increase – aided also by consecutive EU enlargements – up to a point when Eurojust serviced 5 608 cases in 2017 and a further rise to 10 105 cases in 2021. This has meant an 80% increase in just four years (2017-2021), and we look forward to sustaining and supporting the growth curve that lies ahead. However, numbers alone do not fully explain the development of Eurojust and of cross-border judicial cooperation. The countless coordination meetings, coordination centres, conferences and seminars that are conducted at Eurojust not only serve to assist judicial practitioners in their duties, but they also instil a stronger kinship among the community of prosecutors and investigative judges throughout the European Union. This is a qualitative bond that gives us a faster and better understanding of each other’s legal frameworks and requirements. In his article, Judge Bay Larsen, Vice President of the European Court of Justice, offered an excellent summary of the increasing legal complexity that surrounds the principle of mutual recognition in criminal matters and, more specifically, in the execution of European Arrest Warrants. With its in-house legal knowledge and the permanent presence of EU countries’ National Members, Eurojust is much like a hub that gathers Member States’ judicial authorities to support and facilitate work among them. Beyond the numbers and caseload, another qualitative development I would like to highlight is the new Eurojust Regulation of November 2018, which entered into force in December 2019. It updated our governance structure, it transformed us from the EU’s Judicial Cooperation Unit into the EU Agency for Criminal Justice Cooperation, but above all, it offered recognition of the continued need for Eurojust and our role in the European Union’s security architecture. As I write these words, the EU Council and European Parliament have just further extended Eurojust’s mandate with a view to preserving, analysing and storing evidence of core international crimes. While this is largely uncharted territory for Eurojust, we will move fast to live up to these new responsibilities. With armed conflicts continuing in various parts of the world, with the war in Ukraine the latest example, this new strand of work will require our urgent focus. The road ahead promises a number of other developments that will further shape Eurojust as an organisation in years to come. What comes to mind first is the digital shape and form of tomorrow’s criminal justice cooperation, and second, the external dimension of Eurojust’s work. On the first topic, it seems clear that, as our lives increasingly move online, so will criminal justice cooperation. Back in 2018, Eurojust presented to the Council of the European Union the need for a standardised set of digital tools to support cross-border judicial cooperation. In the meantime, the legislative process has been set in motion, and we look forward to a future in which Member States and Eurojust will have equal access to secure e-applications that will define the face of tomorrow’s criminal justice cooperation. The second topic is linked to our work with partner countries outside the European Union. With the advent of Eurojust (2002), the European Arrest Warrant (2002), the European Investigation Order (2014) and the Freezing and Confiscation Order (2018), European judicial cooperation has taken an important turn during the past 20 years. It has also been a time in which Eurojust concluded 13 operational cooperation agreements with a geographically diverse range of countries, from the United States of America to Georgia. In the period that lies ahead, it is our ambition to substantially widen our network with third countries and to welcome additional Liaison Prosecutors to Eurojust to join the 10 colleagues we already host at our premises in The Hague. If the past is any predictor of the future, I imagine Eurojust’s next 20 years will look very different from our experiences so far. With the invaluable support and expertise of policymakers, academics, data protection professionals and, most importantly, the judicial practitioners who decide to place their trust in us, I believe there is important room left to strengthen Eurojust’s role. Our ambition will always be to provide the best possible operational support to prosecutors in the field while remaining at the forefront of major new developments in the field of cross-border judicial cooperation. If we collectively agree to continue on this course, I have no doubt that the words of EU Commissioner Reynders will hold true in the future as much as they do now, when he wrote in his introduction (…) this is also a story of an organisation that has, from the very beginning, been over-delivering. In gratitude to all Eurojust colleagues, past and present, Ladislav Hamran President of Eurojust Ladislav Hamran President of Eurojust
Cross-border cooperation in criminal matters is an important component of European integration and serves to promote respect for #FundamentalRights, says FRA’s @MichaelCJT opening his #Eurojust20 essay.
🐦🔗: https://nitter.eu/EURightsAgency/status/1574344593759502341
Cross-border cooperation in criminal matters is an important component of European integration and serves to promote respect for fundamental rights. Its importance became evident with the progressive elimination of border controls within the European Union, which facilitated free movement within the European Union, but also – at the same time – cross-border crime. The European Union has regulated the main elements of cross-border cooperation in criminal matters. For instance, the European Arrest Warrant (EAW) entering into force in 2004 facilitated the surrender of suspects or convicted persons, without the need to apply long and cumbersome extradition proceedings. The EAW was followed by other framework decisions aiming to facilitate transfer of prisoners, transfer of probation and alternative sanctions, and transfer of pre-trial non-custodial measures. At the same time, the operation of the EAW, which hinges on the principles of mutual trust and mutual recognition, made it clear that the procedural rights guaranteed in national proceedings should be comparable; so courts can easily recognise decisions adopted by other jurisdictions without the need to examine the respect of fundamental rights during the proceedings in another Member State. With the operation of the EAW, it also became clear that conditions of detention (pre-and post-trial) differ in the EU Member States. The Court of Justice of the European Union (CJEU) ruled that Member States executing the EAW should ensure that a requested person would not risk being held in inhumane conditions after the transfer to the issuing Member State. In 2009, the Council of the European Union issued a resolution on a Roadmap for strengthening procedural rights of suspected or accused persons in criminal proceedings. Since then, the EU has put in place six directives on the right to interpretation and translation, information, access to a lawyer, legal aid, the presumption of innocence and the procedural safeguards for children suspected or accused in criminal proceedings. The directives apply to a certain extent also to the cross-border proceedings triggered by the issuing of the EAW. In turn, cross-border collection and sharing of evidence have become increasingly important with the ongoing EU harmonisation in certain fields of criminal law, leading to challenges both in terms of effectiveness as well as safeguarding fundamental rights. This article reflects the work of the European Union Agency for Fundamental Human Rights’ (FRA), which to date has covered three main areas: procedural rights[1], detention conditions, and cross-border collection and exchange of evidence. These areas are particularly pertinent to fundamental rights and important for ensuring the efficiency of cross-border cooperation in criminal matters. Criminal detention and alternatives The EAW framework decision regulating transfer of detained persons between Member States is the most important legal instrument for effective cross-border cooperation in the area of criminal law. The CJEU clarified[2] that Member States executing such transfers must ensure that persons transferred will not run a real risk of inhumane conditions of detention in the requesting Member State. More specifically, in the Dorobantu judgement, the CJEU pointed out that ‘as regards, in particular, the personal space available to each detainee, the executing judicial authority must, in the absence, currently, of minimum standards in that respect under EU law, take account of the minimum requirements under Article 3 of the Convention for the Protection of Human Rights and Fundamental Freedoms’.[3] FRA research showed that the conditions of detention still vary considerably across Member States. Detainees remain vulnerable, as data presented in FRA’s Criminal Detention Database[4] show, with some Member States lacking standards that would meet the minimum requirements set by the European courts. Establishing that such differences exist in the treatment of detainees opens an opportunity for discussing common minimum standards in the EU with respect to pre-trial detention which could improve mutual trust between Member States, increase the effectiveness of mutual recognition instruments, and demonstrate commitment to upholding the EU's fundamental rights and other values. The work of the Agency identified three areas for prioritising action to improve the current situation: first, physical conditions, such as living space, access to sanitary facilities and access to meaningful activities, which should reflect respect for human dignity and other fundamental rights, including privacy and respect for family life. Second, adequate access to healthcare; this requires the presence of a sufficient number of medical staff in detention facilities, which proved to be a particularly important issue during the COVID-19 pandemic. Third, protection from violence. Authorities need to acknowledge that violence is endemic in prison environments and should apply effective measures to protect detainees from violence, both from prison staff and from other inmates. Particular attention should be given to those considered most vulnerable, such as LGBTI prisoners for whom only a few Member States provide special protection measures. One of the underlying reasons for the shortcomings in detention conditions is overcrowding in detention facilities, which is in turn closely related to the overuse of detention measures and underuse of non-custodial measures. In 2016, FRA published a report on fundamental rights aspects of criminal detention and alternatives in EU cross-border transfers[5]. It examined a range of fundamental rights issues related specifically to three EU instruments concerning transfers of prison sentences between EU Member States, probation measures and alternative sanctions, as well as pre-trial supervision measures: the Framework Decision on transfer of prisoners, the Framework Decision on probation and alternative sanctions, and the Framework Decision on the European supervision order. FRA research showed that the EU instruments providing for the cross-border transfer of such alternatives to detention in pre- and post-trial phases remain underutilised, arguably because alternative measures to detention are still not perceived as an effective deterrent to crime. In the light of these findings, FRA continues to encourage Member States to make full use of the framework decisions related to cross-border proceedings, on probation and alternative sanctions and on the European supervision order, but also to use alternative measures more often in domestic proceedings to help reduce prison populations. The use of alternatives is even more important at the pre-trial stage, where individuals remain innocent until proven guilty in a court of law. Promoting and establishing widely alternative measures to detention needs to go hand in hand with a shift in public opinion coupled with a better understanding of their effectiveness and added value to society. At the same time, it would have a clear positive impact both for fundamental rights as well as for further stimulating cross-border cooperation. Criminal procedural rights Over the past years, the FRA has conducted research and published reports concerning the respect of fundamental rights in a number of core areas of criminal procedural rights, covering five directives adopted under the Criminal Procedural Roadmap[6]. The Agency’s research examines the legal provisions in place, as well as the practical implementation of the law to identify shortcomings, as well as promising practices, and suggest improvements. Our research shows, as a general trend, a positive influence of the EU directives on criminal procedural rights. Member States appear to be gradually harmonising the application of EU standards. Nevertheless, challenges persist – for example, information in criminal proceedings is not always conveyed in an understandable manner; access to a lawyer is delayed; and the right to remain silent is not always respected in practice. Moreover, our research shows that these issues continue to affect children too when they are involved in criminal proceedings. Recognising linguistic diversity in the EU, the legislator guaranteed every suspect, accused and requested person the right to interpretation and translation to enable them to participate in proceedings at the same level as persons speaking the language of the proceedings. Our survey of national associations of legal interpreters and translators identified problems particularly with less common languages. For example, respondents in our interviews complained about the inadequate interpretation services and even mentioned using other inmates or family members and friends as interpreters, although poor or inaccurate interpretation could have serious legal consequences. National requirements and practices on interpretation and translation services for criminal proceedings vary, for example as regards the qualifications and certification of official legal interpreters and translators, resulting in varying quality of these services within the EU. The right to information is another important right in criminal proceedings enabling full and engaged participation. The EU has introduced an obligation to provide a written ‘letter of rights’ to those deprived of liberty. However, in practice, as our research indicates, often defendants felt inadequately informed – either because the provision of information is delayed or conveyed as uncomprehensible legal jargon. Access to a lawyer is the most important right in the course of criminal proceedings but also EAW proceedings. Our research found instances where access to a lawyer was delayed and suspects were questioned without the presence of a lawyer. FRA recommended that Member States take appropriate measures to avoid such practices. Additionally, our findings show that authorities do not always inform persons arrested on an EAW about their right to be assisted by lawyers in both states – issuing and executing. Even more so, the authorities do not facilitate this access. Moreover, given the difficulties some defendants deprived of their liberty had in accessing their lawyers, FRA recommended that national authorities should issue specific guidance to law enforcement authorities for prompt, direct and confidential access to a lawyer before the first questioning of defendants deprived of their liberty. FRA provided a range of recommendations for improvement concerning the different criminal procedural rights, some of which have a transnational dimension. For instance, with regard to translation and interpretation, which can be a critical factor for ensuring fair proceedings, we recommend considering cooperation between Member States on interpretation, for example by sharing a pool of interpreters. Moreover, FRA recommends enabling criminal justice authorities to monitor and assess the quality of national interpretation or translation services. The Agency recommended that national authorities put in place robust safeguards to ensure that individuals are effectively informed about their criminal procedural rights as soon as they become suspects. Particular attention should be paid to language barriers, lack of education or any physical or intellectual disability that individuals may have. We also suggested that information should be provided both orally and in writing using non-technical and accessible language. In particular, FRA pointed out that law enforcement should inform any suspect or a potential suspect (sometimes called ‘a person of interest’) about their rights as suspects, in particular the right to remain silent and not to incriminate themselves. With regard to the recurrent practice of questioning as witnesses persons who are likely to become suspects, we call for abandoning this practice to ensure that questioning is immediately stopped once it becomes clear that the person might be charged with a crime, to inform this person fully about their procedural rights, and to enable their consultation with a defence lawyer. Concerning the letter of rights, we recommend introducing a uniform template for all criminal justice authorities in the EU to improve legal certainty and clarity. In addition, as national laws rarely include detailed rules and measures to cater for the needs of persons with disabilities, practical measures, such as transcribing written text into braille for individuals with visual impairments or providing audio files and easy-to-read versions, would improve fundamental rights protection. Specifically concerning children, our ongoing research indicates that when they are involved in criminal proceedings, national authorities do not always follow the safeguards provided by EU law, nor are children treated in the way recommended by the Council of Europe Guidelines on child-friendly justice[7]. We find, for example, that authorities inform children in the same way as adult defendants by handing out a letter of rights or a leaflet; and that their parents are not always involved as required by EU law. On the other hand, we find that children involved in criminal proceedings are rarely detained and, in general, authorities apply non-custodial measures to children. Cross-border collection and sharing of evidence Cross-border collection and sharing of evidence is often necessary for the successful investigation of serious crime, including terrorism. The Agency’s research into the fundamental rights impact of EU counter-terrorism legislation shows that EU level action has helped foster cooperation between Member States when detecting, investigating and prosecuting these offences[8]. Moreover, counter-terrorism practitioners interviewed for this research underlined the added value of Eurojust in setting up and supporting joint investigation teams, in supporting investigations and by exchanging information. The prosecutors and judges interviewed across the EU consider Eurojust as an important facilitator of their work to keep Europe safe. Nevertheless, the research also identified persistent challenges, such as the vague definitions of certain offences which affect legal clarity and foreseeability. Different interpetations of what conduct constitutes an offence such as travelling for the purpose of terrorism, receiving training to terrorism or public provocation to commit a terrorist offence, can have an adverse impact on fundamental rights and discourage lawful conduct but also can hinder cross-border cooperation. Rules for the use of evidence from intelligence work or collected in conflict zones are also not always clear and would benefit from explicit fundamental rights safeguards. Practitioners we interviewed for our research also claimed that information provided by non-EU countries for criminal proceedings in terrorism cases was not systematically verified. It was therefore not clear if it had been obtained legally or not, for example through torture. Such factors can have a real impact on the rights of individuals involved in criminal proceedings and also hinder effective cross-border cooperation in terrorism cases. In closing, I would reiterate the importance of fundamental rights for cross-border cooperatation in criminal matters. FRA research clearly shows that respect for fundamental rights and a clear legislative framework with robust safeguards are prerequisites of Member States’ mutual trust in each other’s justice systems, and therefore for effective cross-border cooperation and the prevention of impunity in the European Union. EU Agencies can assist Member States and EU institutions with capacity building, operational support and data in this important field, and can help promote these shared objectives of criminal justice. Michael O’Flaherty Director of the European Union Agency for Fundamental Rights [1] FRA has undertaken research in this area largely as a result of direct requests from the European Commission. [2] Court of Justice of the European Union, Joined Cases C-404/15 and C-659/15 PPU, Aranyosi and Căldăraru [GC], 5 April 2016 (available on the CJEU’s website). [3] Court of Justice of the European Union, C-128/18, Dorobantu [GC], 15 October 2019 (available on the CJEU’s website). [4] FRA developed the Criminal Detention Database (2015-2019) upon a request by the European Commission as a practical tool to assist members of the judiciary and other legal professionals involved in cross-border criminal proceedings. The database (available on FRA’s website) is a hub for information on detention conditions in all EU Member States. [5] FRA (2016), Criminal detention and alternatives: fundamental rights aspects in EU cross-border transfers (available on FRA’s website). [6] Right to information, Right to interpretation, Right to access to a lawyer, Right to be presumed innocent and to be present of a trial, Safeguards for children in criminal proceedings (available on FRA’s website). [7] Council of Europe (2020), Guidelines of the Committee of Ministers of the Council of Europe on child-friendly justice (available on the CoE’s website). [8] FRA (2021), Directive (EU) 2017/541 on combating terrorism ― Impact on fundamental rights and freedoms (available on FRA’s website).
📘 Today's #Eurojust20 essay:
'The protection of fundamental rights in cross-border cooperation: trends and future perspectives'
- Michael O’Flaherty, Director of the European Union Agency for Fundamental Rights
Cross-border cooperation in criminal matters is an important component of European integration and serves to promote respect for fundamental rights. Its importance became evident with the progressive elimination of border controls within the European Union, which facilitated free movement within the European Union, but also – at the same time – cross-border crime. The European Union has regulated the main elements of cross-border cooperation in criminal matters. For instance, the European Arrest Warrant (EAW) entering into force in 2004 facilitated the surrender of suspects or convicted persons, without the need to apply long and cumbersome extradition proceedings. The EAW was followed by other framework decisions aiming to facilitate transfer of prisoners, transfer of probation and alternative sanctions, and transfer of pre-trial non-custodial measures. At the same time, the operation of the EAW, which hinges on the principles of mutual trust and mutual recognition, made it clear that the procedural rights guaranteed in national proceedings should be comparable; so courts can easily recognise decisions adopted by other jurisdictions without the need to examine the respect of fundamental rights during the proceedings in another Member State. With the operation of the EAW, it also became clear that conditions of detention (pre-and post-trial) differ in the EU Member States. The Court of Justice of the European Union (CJEU) ruled that Member States executing the EAW should ensure that a requested person would not risk being held in inhumane conditions after the transfer to the issuing Member State. In 2009, the Council of the European Union issued a resolution on a Roadmap for strengthening procedural rights of suspected or accused persons in criminal proceedings. Since then, the EU has put in place six directives on the right to interpretation and translation, information, access to a lawyer, legal aid, the presumption of innocence and the procedural safeguards for children suspected or accused in criminal proceedings. The directives apply to a certain extent also to the cross-border proceedings triggered by the issuing of the EAW. In turn, cross-border collection and sharing of evidence have become increasingly important with the ongoing EU harmonisation in certain fields of criminal law, leading to challenges both in terms of effectiveness as well as safeguarding fundamental rights. This article reflects the work of the European Union Agency for Fundamental Human Rights’ (FRA), which to date has covered three main areas: procedural rights[1], detention conditions, and cross-border collection and exchange of evidence. These areas are particularly pertinent to fundamental rights and important for ensuring the efficiency of cross-border cooperation in criminal matters. Criminal detention and alternatives The EAW framework decision regulating transfer of detained persons between Member States is the most important legal instrument for effective cross-border cooperation in the area of criminal law. The CJEU clarified[2] that Member States executing such transfers must ensure that persons transferred will not run a real risk of inhumane conditions of detention in the requesting Member State. More specifically, in the Dorobantu judgement, the CJEU pointed out that ‘as regards, in particular, the personal space available to each detainee, the executing judicial authority must, in the absence, currently, of minimum standards in that respect under EU law, take account of the minimum requirements under Article 3 of the Convention for the Protection of Human Rights and Fundamental Freedoms’.[3] FRA research showed that the conditions of detention still vary considerably across Member States. Detainees remain vulnerable, as data presented in FRA’s Criminal Detention Database[4] show, with some Member States lacking standards that would meet the minimum requirements set by the European courts. Establishing that such differences exist in the treatment of detainees opens an opportunity for discussing common minimum standards in the EU with respect to pre-trial detention which could improve mutual trust between Member States, increase the effectiveness of mutual recognition instruments, and demonstrate commitment to upholding the EU's fundamental rights and other values. The work of the Agency identified three areas for prioritising action to improve the current situation: first, physical conditions, such as living space, access to sanitary facilities and access to meaningful activities, which should reflect respect for human dignity and other fundamental rights, including privacy and respect for family life. Second, adequate access to healthcare; this requires the presence of a sufficient number of medical staff in detention facilities, which proved to be a particularly important issue during the COVID-19 pandemic. Third, protection from violence. Authorities need to acknowledge that violence is endemic in prison environments and should apply effective measures to protect detainees from violence, both from prison staff and from other inmates. Particular attention should be given to those considered most vulnerable, such as LGBTI prisoners for whom only a few Member States provide special protection measures. One of the underlying reasons for the shortcomings in detention conditions is overcrowding in detention facilities, which is in turn closely related to the overuse of detention measures and underuse of non-custodial measures. In 2016, FRA published a report on fundamental rights aspects of criminal detention and alternatives in EU cross-border transfers[5]. It examined a range of fundamental rights issues related specifically to three EU instruments concerning transfers of prison sentences between EU Member States, probation measures and alternative sanctions, as well as pre-trial supervision measures: the Framework Decision on transfer of prisoners, the Framework Decision on probation and alternative sanctions, and the Framework Decision on the European supervision order. FRA research showed that the EU instruments providing for the cross-border transfer of such alternatives to detention in pre- and post-trial phases remain underutilised, arguably because alternative measures to detention are still not perceived as an effective deterrent to crime. In the light of these findings, FRA continues to encourage Member States to make full use of the framework decisions related to cross-border proceedings, on probation and alternative sanctions and on the European supervision order, but also to use alternative measures more often in domestic proceedings to help reduce prison populations. The use of alternatives is even more important at the pre-trial stage, where individuals remain innocent until proven guilty in a court of law. Promoting and establishing widely alternative measures to detention needs to go hand in hand with a shift in public opinion coupled with a better understanding of their effectiveness and added value to society. At the same time, it would have a clear positive impact both for fundamental rights as well as for further stimulating cross-border cooperation. Criminal procedural rights Over the past years, the FRA has conducted research and published reports concerning the respect of fundamental rights in a number of core areas of criminal procedural rights, covering five directives adopted under the Criminal Procedural Roadmap[6]. The Agency’s research examines the legal provisions in place, as well as the practical implementation of the law to identify shortcomings, as well as promising practices, and suggest improvements. Our research shows, as a general trend, a positive influence of the EU directives on criminal procedural rights. Member States appear to be gradually harmonising the application of EU standards. Nevertheless, challenges persist – for example, information in criminal proceedings is not always conveyed in an understandable manner; access to a lawyer is delayed; and the right to remain silent is not always respected in practice. Moreover, our research shows that these issues continue to affect children too when they are involved in criminal proceedings. Recognising linguistic diversity in the EU, the legislator guaranteed every suspect, accused and requested person the right to interpretation and translation to enable them to participate in proceedings at the same level as persons speaking the language of the proceedings. Our survey of national associations of legal interpreters and translators identified problems particularly with less common languages. For example, respondents in our interviews complained about the inadequate interpretation services and even mentioned using other inmates or family members and friends as interpreters, although poor or inaccurate interpretation could have serious legal consequences. National requirements and practices on interpretation and translation services for criminal proceedings vary, for example as regards the qualifications and certification of official legal interpreters and translators, resulting in varying quality of these services within the EU. The right to information is another important right in criminal proceedings enabling full and engaged participation. The EU has introduced an obligation to provide a written ‘letter of rights’ to those deprived of liberty. However, in practice, as our research indicates, often defendants felt inadequately informed – either because the provision of information is delayed or conveyed as uncomprehensible legal jargon. Access to a lawyer is the most important right in the course of criminal proceedings but also EAW proceedings. Our research found instances where access to a lawyer was delayed and suspects were questioned without the presence of a lawyer. FRA recommended that Member States take appropriate measures to avoid such practices. Additionally, our findings show that authorities do not always inform persons arrested on an EAW about their right to be assisted by lawyers in both states – issuing and executing. Even more so, the authorities do not facilitate this access. Moreover, given the difficulties some defendants deprived of their liberty had in accessing their lawyers, FRA recommended that national authorities should issue specific guidance to law enforcement authorities for prompt, direct and confidential access to a lawyer before the first questioning of defendants deprived of their liberty. FRA provided a range of recommendations for improvement concerning the different criminal procedural rights, some of which have a transnational dimension. For instance, with regard to translation and interpretation, which can be a critical factor for ensuring fair proceedings, we recommend considering cooperation between Member States on interpretation, for example by sharing a pool of interpreters. Moreover, FRA recommends enabling criminal justice authorities to monitor and assess the quality of national interpretation or translation services. The Agency recommended that national authorities put in place robust safeguards to ensure that individuals are effectively informed about their criminal procedural rights as soon as they become suspects. Particular attention should be paid to language barriers, lack of education or any physical or intellectual disability that individuals may have. We also suggested that information should be provided both orally and in writing using non-technical and accessible language. In particular, FRA pointed out that law enforcement should inform any suspect or a potential suspect (sometimes called ‘a person of interest’) about their rights as suspects, in particular the right to remain silent and not to incriminate themselves. With regard to the recurrent practice of questioning as witnesses persons who are likely to become suspects, we call for abandoning this practice to ensure that questioning is immediately stopped once it becomes clear that the person might be charged with a crime, to inform this person fully about their procedural rights, and to enable their consultation with a defence lawyer. Concerning the letter of rights, we recommend introducing a uniform template for all criminal justice authorities in the EU to improve legal certainty and clarity. In addition, as national laws rarely include detailed rules and measures to cater for the needs of persons with disabilities, practical measures, such as transcribing written text into braille for individuals with visual impairments or providing audio files and easy-to-read versions, would improve fundamental rights protection. Specifically concerning children, our ongoing research indicates that when they are involved in criminal proceedings, national authorities do not always follow the safeguards provided by EU law, nor are children treated in the way recommended by the Council of Europe Guidelines on child-friendly justice[7]. We find, for example, that authorities inform children in the same way as adult defendants by handing out a letter of rights or a leaflet; and that their parents are not always involved as required by EU law. On the other hand, we find that children involved in criminal proceedings are rarely detained and, in general, authorities apply non-custodial measures to children. Cross-border collection and sharing of evidence Cross-border collection and sharing of evidence is often necessary for the successful investigation of serious crime, including terrorism. The Agency’s research into the fundamental rights impact of EU counter-terrorism legislation shows that EU level action has helped foster cooperation between Member States when detecting, investigating and prosecuting these offences[8]. Moreover, counter-terrorism practitioners interviewed for this research underlined the added value of Eurojust in setting up and supporting joint investigation teams, in supporting investigations and by exchanging information. The prosecutors and judges interviewed across the EU consider Eurojust as an important facilitator of their work to keep Europe safe. Nevertheless, the research also identified persistent challenges, such as the vague definitions of certain offences which affect legal clarity and foreseeability. Different interpetations of what conduct constitutes an offence such as travelling for the purpose of terrorism, receiving training to terrorism or public provocation to commit a terrorist offence, can have an adverse impact on fundamental rights and discourage lawful conduct but also can hinder cross-border cooperation. Rules for the use of evidence from intelligence work or collected in conflict zones are also not always clear and would benefit from explicit fundamental rights safeguards. Practitioners we interviewed for our research also claimed that information provided by non-EU countries for criminal proceedings in terrorism cases was not systematically verified. It was therefore not clear if it had been obtained legally or not, for example through torture. Such factors can have a real impact on the rights of individuals involved in criminal proceedings and also hinder effective cross-border cooperation in terrorism cases. In closing, I would reiterate the importance of fundamental rights for cross-border cooperatation in criminal matters. FRA research clearly shows that respect for fundamental rights and a clear legislative framework with robust safeguards are prerequisites of Member States’ mutual trust in each other’s justice systems, and therefore for effective cross-border cooperation and the prevention of impunity in the European Union. EU Agencies can assist Member States and EU institutions with capacity building, operational support and data in this important field, and can help promote these shared objectives of criminal justice. Michael O’Flaherty Director of the European Union Agency for Fundamental Rights [1] FRA has undertaken research in this area largely as a result of direct requests from the European Commission. [2] Court of Justice of the European Union, Joined Cases C-404/15 and C-659/15 PPU, Aranyosi and Căldăraru [GC], 5 April 2016 (available on the CJEU’s website). [3] Court of Justice of the European Union, C-128/18, Dorobantu [GC], 15 October 2019 (available on the CJEU’s website). [4] FRA developed the Criminal Detention Database (2015-2019) upon a request by the European Commission as a practical tool to assist members of the judiciary and other legal professionals involved in cross-border criminal proceedings. The database (available on FRA’s website) is a hub for information on detention conditions in all EU Member States. [5] FRA (2016), Criminal detention and alternatives: fundamental rights aspects in EU cross-border transfers (available on FRA’s website). [6] Right to information, Right to interpretation, Right to access to a lawyer, Right to be presumed innocent and to be present of a trial, Safeguards for children in criminal proceedings (available on FRA’s website). [7] Council of Europe (2020), Guidelines of the Committee of Ministers of the Council of Europe on child-friendly justice (available on the CoE’s website). [8] FRA (2021), Directive (EU) 2017/541 on combating terrorism ― Impact on fundamental rights and freedoms (available on FRA’s website).
RT @Eurojust: 📘 Today's #Eurojust20 essay:
'AI and data protection in judicial cooperation in criminal matters'
- Wojciech Wiewiórowski, European Data Protection Supervisor, and Michał Fila, Legal Officer at the European Data Protection Supervisor
Introduction As we celebrate the 20th Anniversary of Eurojust, the European body for judicial cooperation in criminal matters, we should also review the historical and technological contexts in which the agency began its existence. The year is 2002: the European Union consists of 14 Member States; new Euro banknotes and coins are in our pockets; the future of the Nice Treaty in the period between Irish referendums is uncertain; the role of the year-old Charter of Fundamental Rights is uncertain as well; and expectations are high for the newly established European Convention led by Valéry Giscard d’Estaing, who was in charge of writing the Constitution for Europe. We were also just four months on from 9/11. Internet Explorer was occupying more than 90% of the market; Safari and Firefox did not yet exist. Using your European mobile phone (smartphones were not yet known) in the United States or Japan was difficult and barely affordable. The IT market was recovering after the dot.com bubble collapse. What were we, the authors, doing 20 years ago? In 2002, Wojciech was teaching constitutional and European law as a young PhD student, exploring the interplay between IT and law, both academically and professionally. Michał was defending his Master of Laws (LL.M) thesis on police cooperation in Europe at the Christian Albrecht University in Kiel, Germany. Our interest in the newly created Eurojust was limited. Although some specific and focused solutions, developed by artificial intelligence (AI) researchers, were being widely used at that time, they were still only rarely described as ‘artificial intelligence’. The only remote association made between the judiciary and AI in popular culture was probably through the figure of Judge Dredd! Fast forward to 2022 and here we are, with national strategies, policies and regulations on AI adopted by almost all major economies in the world. Non-binding guidelines or principles for the use of AI, focusing on ethical considerations, are common. Proposals for legal changes to address issues raised by AI (for example, transparency) are tabled in the European Union, the United Kingdom, the United States and around the globe. At the same time, Wojciech is at the helm of the EU’s supervisory authority responsible for monitoring compliance with data protection rules by all EU institutions, offices, bodies and agencies (EUIs), including, since December 2019, Eurojust. Since September 2020, Michał has been the legal officer at the European Data Protection Supervisor (EDPS) responsible for relations with Eurojust. The EDPS took over the supervision of Eurojust at a crucial time – in 2020, the European Commission (EC) presented its Communication on the Digitalisation of Justice in the European Union[1]. One of the objectives set out in the EC’s document is to further improve cross-border judicial cooperation between competent authorities at the European level. To this end, the EC announced that it is exploring ways to increase the availability of relevant machine-readable data produced by the judiciary, in order to establish trustworthy machine-learning AI solutions for interested stakeholders to use. Shortly after, in April 2021, the EC presented a proposal for an AI Regulation laying down harmonised rules for the EU, otherwise known as the Artificial Intelligence Act (AI Act)[2]. In both of these contexts, the EC stressed that any actions put in place must be in full compliance with the EU’s fundamental rights, including the right to the protection of personal data. The AI Act would also designate the EDPS as the competent authority for the supervision of EUIs as they develop and use AI systems[3]. The use of AI tools in the area of justice may represent a high risk to the fundamental rights of individuals[4]. This is especially true with regard to AI systems that may be used to assist judicial authorities in factual and legal research, as well as in interpreting and applying the results of such research in a specific case. Such high risk is largely absent in cases where AI systems are used for purely ancillary administrative activities that do not affect the actual administration of justice in individual cases, such as anonymisation/pseudonymisation of judicial decisions/documents or purely administrative tasks and allocation of resources. The formal views of the EDPS and of the European Data Protection Board (EDPB) on the new regulatory framework are expressed in their joint opinion issued in June 2021[5]. With this written contribution for Eurojust’s 20th anniversary, we take this opportunity to reflect on some of the data protection issues stemming from the proposed AI rules on one hand, and the ongoing reform of Eurojust on the other. Relationship between the data protection framework and AI rules When speaking about AI, we usually start by reminding readers that a comprehensive European data protection framework, adopted on the basis of Article 16 TFEU, already exists. The data protection framework of Eurojust consists of the Data Protection Regulation for the EUIs (EUDPR)[6] and the specifying data protection provisions of the Eurojust Regulation.[7] While the Law Enforcement Directive (LED)[8] is not directly applicable to Eurojust, it determines the way in which national judicial authorities of Member States protect personal data for the purposes of prevention, investigation, detection and prosecution of criminal offences or the execution of criminal penalties. Contrary to the European Public Prosecutor’s Office and Europol, to which the EUDPR does not apply for the processing of operational personal data[9], the Eurojust data protection framework can be regarded as both clearer and more comprehensive. The EUDPR governs the processing of administrative personal data, and, together, its Chapter IX and the provisions of the Eurojust Regulation, constituting a lex specialis to the general rules, apply to the processing of operational personal data. We must stress the need for consistent interpretation and application of these rules – something that the text itself underlines[10]. It should be clearly stated that the existing data protection rules apply to the processing of personal data by Eurojust, whenever carried out wholly or partly by automated means, including possible processing by AI systems. There should be no doubt that the essential data protection requirements, derived from Article 8 of the EU Charter of Fundamental Rights, such as the principles of necessity, proportionality, accuracy, purpose limitation, data minimisation, integrity and confidentiality, continue to apply. Other obligations of the controller, such as data protection by design and by default, are also relevant. Whenever personal data is processed, data protection provisions apply. It should be clear that, when it comes to the processing of personal data, the new AI regulation would be without prejudice to the existing rules[11]. Human involvement One of these rules deserves a special mention here. Article 77 of the EUDPR prohibits a ‘decision based solely on automated processing’, unless authorised by EU law as providing adequate safeguards – which should include at least the right to obtain human intervention from the controller. Such decisions (if authorised by law) shall not be based on sensitive data ‘unless suitable measures to safeguard the data subjects’ rights, freedom and legitimate interests are in place’. There is a clear requirement for specific safeguards to tackle the risks linked to the processing of sensitive data in automated processing used for decision-making. To that end, controllers need to provide for human involvement in the processes where AI operates. The use of AI systems should involve systematic human intervention, evaluation and validation by expert staff. Human validation should be employed as an inherent step to ensure that the output of the systems is faultless. In the case that the automated results are assessed as faulty, the human intervention should provide feedback to be recorded and used for retraining the AI. How to best implement meaningful human involvement is certainly a topic for another article; here, we want only to stress the importance of such a safeguard, while being mindful that it is not the only factor to consider. AI training and data minimisation AI regulation discourse often seems to avoid the problem of potential conflict between AI development and the data minimisation principle. According to the conventional understanding of AI, data is an essential strategic resource and any meaningful progress in cutting-edge AI techniques requires large volumes of data, including personal data. Training of AI models relies on the data ‘feeding’ them. The more and better-quality data used, the better the AI tool is trained. AI developers are constantly seeking datasets that could improve the functioning of their creations. However, such an approach is in opposition to the principle of data minimisation. This fundamental principle is rarely considered when discussing AI regulation. However, it remains applicable to any processing of personal data. Designers and developers should therefore ask themselves whether it is really necessary to train a particular model on personal data. The data minimisation principle, combined with the principles of data protection by design and by default, are general requirements when using anonymous data if possible[12]. If the AI tool can be trained on anonymised datasets, collecting or injecting personal data in the training process should not take place. Current research demonstrates that AI is not synonymous with big data, and there are several other approaches that can be used in different small data settings[13]. Is AI a silver bullet? We all know that digital transformation has profoundly changed people’s lives in recent decades and will continue to do so. The use of AI in the public sector, including in the area of criminal justice and cross-border cooperation, is increasingly being explored. We understand there are high expectations regarding the possible benefits of these solutions; for instance, to help make judicial decisions machine readable, to simplify the reuse of case-law or simply to improve legal practitioners’ advice to clients. Although AI can be used in process automation, it should not be seen as a universal solution to all problems and shortcomings. Even when the development of AI is delegated to a third party, the process of correctly developing an AI system demands the work and attention of people who know how the organisation works. It is a fallacy to believe that AI will, by itself, magically correct procedures that were already problematic. While digital tools often contribute to the greater efficiency and effectiveness of today’s judicial systems, it is crucial that their deployment should take into account the requirements to guarantee higher standards for the public justice service as well as the expectations and needs of the justice system’s professionals and users. The use of digital technologies in the justice sector is highly sensitive and must therefore meet state-of-the-art standards with regard to information security and cyber security, and must fully comply with privacy and data protection legislation and with the standards upheld by the rule of law. When discussing the use cases of AI models with representatives of law enforcement and of the judiciary, we are often given the impression that the principles of necessity and proportionality in particular are not sufficiently addressed. We believe that the development of machine-learning models needs to be driven by the proven ability of the model to fulfil a specific and legitimate purpose and not by the availability of the technology. In assessing necessity, EU entities should demonstrate that their purposes could not be accomplished in another reasonable way[14]. They should demonstrate a real need for AI to process personal data, how the processing effectively addresses this need and that the same purpose cannot be reasonably achieved with other, less invasive means. The main argument made in this context is that the growing volume of processed datasets can indeed be considered a starting point for the necessity assessment. This argument may provide a general reason for the use of AI to effectively carry out specific tasks entrusted to EUIs. Nevertheless, there are still elements that need to be added in order to complete the necessity assessment. Such assessment should explain and document why some AI models are preferred to others, to justify the selection of the least intrusive solution from a personal data perspective. Possible use cases of AI systems for Eurojust Given Eurojust’s role as the EU hub for supporting and strengthening judicial cooperation between national authorities in charge of investigating and prosecuting serious crime, it seems that certain types of AI applications would fit this role better than others. For example, if we consider Eurojust as an agency that does not conduct its own investigations, tools for forensic analysis or visual biometric identification would not be at the top of the list, especially given the strong reservations around the intrusiveness of such means and the potential overlap with other actors, such as Europol. However, there are other AI categories that seem highly relevant for cross-border judicial cooperation, such as various natural language processing (NLP) tools. These technologies are particularly useful for the processing of large-scale sets of unstructured data, commonly handled by judicial authorities. NLP technologies can support and facilitate Eurojust’s main tasks by improving its internal processes; for example, these tools can be used for automated document processing, machine translation in cross-border cases, text summarisation or named-entity recognition. Automated document processing Considering that Eurojust is starting the process of designing and developing its new case management system, automated document processing (ADP) seems an obvious candidate for a use case[15]. ADP proves to be particularly valuable for processing high volumes of documents, especially for the classification, conversion and archiving of these documents in searchable formats. These types of AI systems can not only significantly reduce the need for manual document processing, but can also contribute to improving data accuracy and completeness. The conversion of paper-based formats into searchable documents is also the first step in exploring further deployment of other AI-driven tools, such as machine translation. Automated translation Overcoming language and communication difficulties between judicial authorities of the EU Member States was one of the driving forces behind the creation of Eurojust. It is also a strong argument for the application of AI in the context of cross-border cooperation in criminal justice. The need to communicate and analyse evidence in multiple languages is self-explanatory, particularly for joint investigative teams (JITs) supported by Eurojust[16]. Integrating automated translation tools into JITs’ operations could significantly reduce the time spent on translation and make the evidence directly accessible to all team members; not to mention the reduction in costs for sworn translation, which would still be necessary for evidence to be admissible in court. However, the specificity of cross-border judicial cooperation seems to be a problem when it comes to machine translation. Domain-specific legal language can pose a challenge to generic automated translation systems available on the market, as they are not reliable when distinguishing specific legal terminology from the generic language. To produce a high-quality translation, domain-specific terminology needs to be ‘learned’ and integrated into the AI tool. The research in this area is advanced and has generated promising results[17]. Nevertheless, domain-specific customisation would still require time and significant resources. Automated summarisation systems Another type of NLP tool to support cross-border criminal justice cooperation is text summarisation (summarisation systems). These tools prove to be particularly useful in applications where large amounts of information need to be processed in a limited amount of time. Summarisation systems facilitate the extracting of the most relevant information, significantly reducing the time needed to analyse large volumes of text, such as documentation seized in criminal investigations. Summarisation systems can also improve data classification and accessibility, especially in cases where processing by humans would take too long and where precision is not decisive. Legal research We turn now to another use case for NLP technologies: their use in legal research to facilitate the identification of case-relevant statutes, provisions and case-law. While this might be dispensable for research on the law of the EU Member States or non-EU countries posting Liaison Prosecutors to Eurojust (with Eurojust here fulfilling its role as a knowledge hub), there are instances where knowledge of foreign law is necessary for Eurojust to make informed decisions concerning data protection. We refer to the assessments of appropriate safeguards, provided for in Article 56 of the Eurojust Regulation. Knowledge about foreign data protection regulations applicable in the transfer of operational personal data to non-EU countries is an important element of Eurojust’s assessment of existing data protection safeguards. This is a potential use case where AI technology could directly support the application of data protection provisions. Moreover, legal research supported by AI would not require the AI tool to process individuals’ personal data. However, linguistic barriers might be a particular challenge in these situations, making this another case where automated translation could come in handy. The AI Act and Eurojust’s cooperation with third countries Since we have already mentioned Eurojust transfers to third countries, allow us another digression on this point. Some of the solutions proposed by the AI Act might appear complicated when it comes to Eurojust’s relations with external partners. The EC proposes to limit the scope of the AI Act with regard to international law enforcement and judicial cooperation. This would mean that the provisions of the draft AI Act, according to its Article 2(4), would not apply to public authorities in a third country or to international organisations, if these authorities or organisations use AI systems in the framework of international agreements for law enforcement and judicial cooperation with the EU, or with one or more EU Member State. In our view, Article 2(4) would not, in any way, limit the application of the AI Act to EUIs; only public authorities in third countries and relevant international organisations could ‘benefit’ from this proposed exception. The practical application of such an exception in the Eurojust environment raises some questions. While the AI Act would be applicable to Eurojust as it develops or uses AI systems, does this exemption mean that it would not (formally) be applicable to third countries’ Liaison Prosecutors operating at Eurojust? We feel this issue merits some further reflection in advance of the negotiations between legislators. Prior assessment and data protection by design and by default While the EDPS takes note of new and emerging ideas, it is not our intention, nor our role, to plead for these ideas to be put in place. Prior to the set-up of these technologies, authorities considering such applications should perform a legal and ethical assessment to take into account the impact and any possible risk to the fundamental rights and freedoms of individuals, as well as their ethical and legal implications. It is also important to conduct the testing and evaluation of these technologies to ensure that their performance meet the relevant standards, especially regarding data accuracy and bias. The processing of personal data is often at the heart of AI technologies. At the same time, the data collected, processed and stored in judicial systems may be highly sensitive, revealing intimate details about individuals or even causing a threat to their lives. Giving access to this data for the purpose of training algorithms has to be considered with extreme caution and under very strict conditions. Training, testing and validation of machine-learning models with operational personal data and for their further use in the context of a specific Eurojust activity should not be carried out before a data protection impact assessment is done, according to Article 89 of the EUDPR. In addition, we stress that the responsibility of the controller goes beyond that: it starts with adequate project governance, which should take into account the principle of data protection by design throughout the conception and development of the AI tool and system in question. A data-protection compliant AI tool or system can be achieved once the following are in place: clear commitments to this principle in the key documents of the project; policies, processes and methodologies that consider data protection at each stage of the project; by identifying privacy and data protection stakeholders; by assigning roles and responsibilities regarding data protection; by working with competent individuals; and by properly documenting all of these steps. Furthermore, sets of business-level requirements on data protection and mechanisms to assess compliance of the outcome are needed. The controller also needs to put in place procedures for the identification and elimination of any bias in the data used to further train AI models, and to verify that the training data used does not cause discrimination. Processes to check the training or validation of data sets must be built and documented, and procedures allowing for regular monitoring of the models regarding biases and their readjustment or retraining must exist. These processes should include statistical checks on the input and output data. Final remarks From an EDPS perspective, we can clearly see the added value of AI. AI solutions can help complete tasks in a much faster and more cost-effective way, and can also be more accurate and precise than humans, if deployed correctly. At a time when nearly all judicial systems are facing a backlog of cases to process, the promises of efficiency that AI brings cannot be ignored. AI can also detect duplicated information in a reliable way, which contributes to data minimisation and helps to reduce personal data processing by effective anonymisation. If correctly put in place, AI may help to reach true equality and improve access to impartial and objective justice. Nevertheless, we also see the associated risks. Algorithms are only as good as their programmers and the data they have been trained on. This leaves AI systems vulnerable to human error or historical bias. Gains in speed and efficiency can easily turn into disadvantages, if personal data is collected and processed in an immanently biased way. Lack of human oversight and monitoring mechanisms may have dire consequences for the fundamental rights of individuals, as well as their trust in judicial systems and in the EU mechanisms supporting them. Finally, we see many actors in the field trying to be the first to seize the potential benefits of AI. There is a need for a coordinated approach at EU level when it comes to EUIs’ development and use of AI systems to support law enforcement and judicial cooperation. You can count on the EDPS to play its part in the EU’s coordinated approach to AI. Wojciech Wiewiórowski European Data Protection SupervisorMichał Fila Legal Officer at the European Data Protection Supervisor [1] COM(2020) 710 final; Communication of 2 December 2020 from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions ‘Digitalisation of justice in the European Union – a toolbox of opportunities’. [2] COM(2021) 206 final; Proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (Artificial Intelligence Act) and amending certain union legislative acts. [3] Article 2 and 3 of the draft AI Act. [4] The proposed AI Act would explicitly qualify ‘AI systems intended to assist a judicial authority in researching and interpreting facts and the law and in applying the law to a concrete set of facts’ as a high-risk system subject to a particular legal regime (see Annex III, point 8 (a)). [5] EDPB-EDPS Joint Opinion 5/2021 on the proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (Artificial Intelligence Act), 18 June 2021 (available on the EDPS’s website). [6] Regulation (EU) 2018/1725 of the European Parliament and of the Council of 23 October 2018 on the protection of natural persons with regard to the processing of personal data by the Union institutions, bodies, offices and agencies and on the free movement of such data, and repealing Regulation (EC) No 45/2001 and Decision No 1247/2002/EC, OJ L 295, 21.11.2018, pp. 39–98. [7] In particular, Chapter IV of the Eurojust Regulation. [8] Directive (EU) 2016/680 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data by competent authorities for the purposes of the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, and on the free movement of such data, and repealing Council Framework Decision 2008/977/JHA, OJ L 119, 4.5.2016, pp. 89–131. [9] Article 2(2) and 2(3) of the Regulation 2018/1725. [10] Recital 29 of the Eurojust Regulation. [11] With some notable exceptions, such as the list of prohibited AI practices in Article 5 of the AI Act. [12] For more details on data protection by design as an enforceable legal obligation, see the EDPS Preliminary Opinion no. 5/2018 on Privacy by Design (available on the EDPS’s website). [13] Small Data’s Big AI Potential (available on the Center for Security and Emerging Technology’s website). [14] See also the EDPS Toolkit on Assessing the necessity of measures that limit the fundamental right to the protection of personal data, as well as the EDPS quick guide to necessity and proportionality (available on the EDPS’s website) [15] Cross-border Digital Criminal Justice, Final Report (pp. 160-162) (available on the Publications Office of the European Union’s website). [16] COM(2021) 756 final; Analytical supporting document accompanying the proposal for the Regulation on the JIT collaboration platform (available on the European Commission’s website) [17] See for instance the Connecting Europe Facility Digital programme (CEF Digital) with the eTranslation tool (available on the European Commission’s website).
📘 Today's #Eurojust20 essay:
'AI and data protection in judicial cooperation in criminal matters'
- Wojciech Wiewiórowski, European Data Protection Supervisor, and Michał Fila, Legal Officer at the European Data Protection Supervisor
Introduction As we celebrate the 20th Anniversary of Eurojust, the European body for judicial cooperation in criminal matters, we should also review the historical and technological contexts in which the agency began its existence. The year is 2002: the European Union consists of 14 Member States; new Euro banknotes and coins are in our pockets; the future of the Nice Treaty in the period between Irish referendums is uncertain; the role of the year-old Charter of Fundamental Rights is uncertain as well; and expectations are high for the newly established European Convention led by Valéry Giscard d’Estaing, who was in charge of writing the Constitution for Europe. We were also just four months on from 9/11. Internet Explorer was occupying more than 90% of the market; Safari and Firefox did not yet exist. Using your European mobile phone (smartphones were not yet known) in the United States or Japan was difficult and barely affordable. The IT market was recovering after the dot.com bubble collapse. What were we, the authors, doing 20 years ago? In 2002, Wojciech was teaching constitutional and European law as a young PhD student, exploring the interplay between IT and law, both academically and professionally. Michał was defending his Master of Laws (LL.M) thesis on police cooperation in Europe at the Christian Albrecht University in Kiel, Germany. Our interest in the newly created Eurojust was limited. Although some specific and focused solutions, developed by artificial intelligence (AI) researchers, were being widely used at that time, they were still only rarely described as ‘artificial intelligence’. The only remote association made between the judiciary and AI in popular culture was probably through the figure of Judge Dredd! Fast forward to 2022 and here we are, with national strategies, policies and regulations on AI adopted by almost all major economies in the world. Non-binding guidelines or principles for the use of AI, focusing on ethical considerations, are common. Proposals for legal changes to address issues raised by AI (for example, transparency) are tabled in the European Union, the United Kingdom, the United States and around the globe. At the same time, Wojciech is at the helm of the EU’s supervisory authority responsible for monitoring compliance with data protection rules by all EU institutions, offices, bodies and agencies (EUIs), including, since December 2019, Eurojust. Since September 2020, Michał has been the legal officer at the European Data Protection Supervisor (EDPS) responsible for relations with Eurojust. The EDPS took over the supervision of Eurojust at a crucial time – in 2020, the European Commission (EC) presented its Communication on the Digitalisation of Justice in the European Union[1]. One of the objectives set out in the EC’s document is to further improve cross-border judicial cooperation between competent authorities at the European level. To this end, the EC announced that it is exploring ways to increase the availability of relevant machine-readable data produced by the judiciary, in order to establish trustworthy machine-learning AI solutions for interested stakeholders to use. Shortly after, in April 2021, the EC presented a proposal for an AI Regulation laying down harmonised rules for the EU, otherwise known as the Artificial Intelligence Act (AI Act)[2]. In both of these contexts, the EC stressed that any actions put in place must be in full compliance with the EU’s fundamental rights, including the right to the protection of personal data. The AI Act would also designate the EDPS as the competent authority for the supervision of EUIs as they develop and use AI systems[3]. The use of AI tools in the area of justice may represent a high risk to the fundamental rights of individuals[4]. This is especially true with regard to AI systems that may be used to assist judicial authorities in factual and legal research, as well as in interpreting and applying the results of such research in a specific case. Such high risk is largely absent in cases where AI systems are used for purely ancillary administrative activities that do not affect the actual administration of justice in individual cases, such as anonymisation/pseudonymisation of judicial decisions/documents or purely administrative tasks and allocation of resources. The formal views of the EDPS and of the European Data Protection Board (EDPB) on the new regulatory framework are expressed in their joint opinion issued in June 2021[5]. With this written contribution for Eurojust’s 20th anniversary, we take this opportunity to reflect on some of the data protection issues stemming from the proposed AI rules on one hand, and the ongoing reform of Eurojust on the other. Relationship between the data protection framework and AI rules When speaking about AI, we usually start by reminding readers that a comprehensive European data protection framework, adopted on the basis of Article 16 TFEU, already exists. The data protection framework of Eurojust consists of the Data Protection Regulation for the EUIs (EUDPR)[6] and the specifying data protection provisions of the Eurojust Regulation.[7] While the Law Enforcement Directive (LED)[8] is not directly applicable to Eurojust, it determines the way in which national judicial authorities of Member States protect personal data for the purposes of prevention, investigation, detection and prosecution of criminal offences or the execution of criminal penalties. Contrary to the European Public Prosecutor’s Office and Europol, to which the EUDPR does not apply for the processing of operational personal data[9], the Eurojust data protection framework can be regarded as both clearer and more comprehensive. The EUDPR governs the processing of administrative personal data, and, together, its Chapter IX and the provisions of the Eurojust Regulation, constituting a lex specialis to the general rules, apply to the processing of operational personal data. We must stress the need for consistent interpretation and application of these rules – something that the text itself underlines[10]. It should be clearly stated that the existing data protection rules apply to the processing of personal data by Eurojust, whenever carried out wholly or partly by automated means, including possible processing by AI systems. There should be no doubt that the essential data protection requirements, derived from Article 8 of the EU Charter of Fundamental Rights, such as the principles of necessity, proportionality, accuracy, purpose limitation, data minimisation, integrity and confidentiality, continue to apply. Other obligations of the controller, such as data protection by design and by default, are also relevant. Whenever personal data is processed, data protection provisions apply. It should be clear that, when it comes to the processing of personal data, the new AI regulation would be without prejudice to the existing rules[11]. Human involvement One of these rules deserves a special mention here. Article 77 of the EUDPR prohibits a ‘decision based solely on automated processing’, unless authorised by EU law as providing adequate safeguards – which should include at least the right to obtain human intervention from the controller. Such decisions (if authorised by law) shall not be based on sensitive data ‘unless suitable measures to safeguard the data subjects’ rights, freedom and legitimate interests are in place’. There is a clear requirement for specific safeguards to tackle the risks linked to the processing of sensitive data in automated processing used for decision-making. To that end, controllers need to provide for human involvement in the processes where AI operates. The use of AI systems should involve systematic human intervention, evaluation and validation by expert staff. Human validation should be employed as an inherent step to ensure that the output of the systems is faultless. In the case that the automated results are assessed as faulty, the human intervention should provide feedback to be recorded and used for retraining the AI. How to best implement meaningful human involvement is certainly a topic for another article; here, we want only to stress the importance of such a safeguard, while being mindful that it is not the only factor to consider. AI training and data minimisation AI regulation discourse often seems to avoid the problem of potential conflict between AI development and the data minimisation principle. According to the conventional understanding of AI, data is an essential strategic resource and any meaningful progress in cutting-edge AI techniques requires large volumes of data, including personal data. Training of AI models relies on the data ‘feeding’ them. The more and better-quality data used, the better the AI tool is trained. AI developers are constantly seeking datasets that could improve the functioning of their creations. However, such an approach is in opposition to the principle of data minimisation. This fundamental principle is rarely considered when discussing AI regulation. However, it remains applicable to any processing of personal data. Designers and developers should therefore ask themselves whether it is really necessary to train a particular model on personal data. The data minimisation principle, combined with the principles of data protection by design and by default, are general requirements when using anonymous data if possible[12]. If the AI tool can be trained on anonymised datasets, collecting or injecting personal data in the training process should not take place. Current research demonstrates that AI is not synonymous with big data, and there are several other approaches that can be used in different small data settings[13]. Is AI a silver bullet? We all know that digital transformation has profoundly changed people’s lives in recent decades and will continue to do so. The use of AI in the public sector, including in the area of criminal justice and cross-border cooperation, is increasingly being explored. We understand there are high expectations regarding the possible benefits of these solutions; for instance, to help make judicial decisions machine readable, to simplify the reuse of case-law or simply to improve legal practitioners’ advice to clients. Although AI can be used in process automation, it should not be seen as a universal solution to all problems and shortcomings. Even when the development of AI is delegated to a third party, the process of correctly developing an AI system demands the work and attention of people who know how the organisation works. It is a fallacy to believe that AI will, by itself, magically correct procedures that were already problematic. While digital tools often contribute to the greater efficiency and effectiveness of today’s judicial systems, it is crucial that their deployment should take into account the requirements to guarantee higher standards for the public justice service as well as the expectations and needs of the justice system’s professionals and users. The use of digital technologies in the justice sector is highly sensitive and must therefore meet state-of-the-art standards with regard to information security and cyber security, and must fully comply with privacy and data protection legislation and with the standards upheld by the rule of law. When discussing the use cases of AI models with representatives of law enforcement and of the judiciary, we are often given the impression that the principles of necessity and proportionality in particular are not sufficiently addressed. We believe that the development of machine-learning models needs to be driven by the proven ability of the model to fulfil a specific and legitimate purpose and not by the availability of the technology. In assessing necessity, EU entities should demonstrate that their purposes could not be accomplished in another reasonable way[14]. They should demonstrate a real need for AI to process personal data, how the processing effectively addresses this need and that the same purpose cannot be reasonably achieved with other, less invasive means. The main argument made in this context is that the growing volume of processed datasets can indeed be considered a starting point for the necessity assessment. This argument may provide a general reason for the use of AI to effectively carry out specific tasks entrusted to EUIs. Nevertheless, there are still elements that need to be added in order to complete the necessity assessment. Such assessment should explain and document why some AI models are preferred to others, to justify the selection of the least intrusive solution from a personal data perspective. Possible use cases of AI systems for Eurojust Given Eurojust’s role as the EU hub for supporting and strengthening judicial cooperation between national authorities in charge of investigating and prosecuting serious crime, it seems that certain types of AI applications would fit this role better than others. For example, if we consider Eurojust as an agency that does not conduct its own investigations, tools for forensic analysis or visual biometric identification would not be at the top of the list, especially given the strong reservations around the intrusiveness of such means and the potential overlap with other actors, such as Europol. However, there are other AI categories that seem highly relevant for cross-border judicial cooperation, such as various natural language processing (NLP) tools. These technologies are particularly useful for the processing of large-scale sets of unstructured data, commonly handled by judicial authorities. NLP technologies can support and facilitate Eurojust’s main tasks by improving its internal processes; for example, these tools can be used for automated document processing, machine translation in cross-border cases, text summarisation or named-entity recognition. Automated document processing Considering that Eurojust is starting the process of designing and developing its new case management system, automated document processing (ADP) seems an obvious candidate for a use case[15]. ADP proves to be particularly valuable for processing high volumes of documents, especially for the classification, conversion and archiving of these documents in searchable formats. These types of AI systems can not only significantly reduce the need for manual document processing, but can also contribute to improving data accuracy and completeness. The conversion of paper-based formats into searchable documents is also the first step in exploring further deployment of other AI-driven tools, such as machine translation. Automated translation Overcoming language and communication difficulties between judicial authorities of the EU Member States was one of the driving forces behind the creation of Eurojust. It is also a strong argument for the application of AI in the context of cross-border cooperation in criminal justice. The need to communicate and analyse evidence in multiple languages is self-explanatory, particularly for joint investigative teams (JITs) supported by Eurojust[16]. Integrating automated translation tools into JITs’ operations could significantly reduce the time spent on translation and make the evidence directly accessible to all team members; not to mention the reduction in costs for sworn translation, which would still be necessary for evidence to be admissible in court. However, the specificity of cross-border judicial cooperation seems to be a problem when it comes to machine translation. Domain-specific legal language can pose a challenge to generic automated translation systems available on the market, as they are not reliable when distinguishing specific legal terminology from the generic language. To produce a high-quality translation, domain-specific terminology needs to be ‘learned’ and integrated into the AI tool. The research in this area is advanced and has generated promising results[17]. Nevertheless, domain-specific customisation would still require time and significant resources. Automated summarisation systems Another type of NLP tool to support cross-border criminal justice cooperation is text summarisation (summarisation systems). These tools prove to be particularly useful in applications where large amounts of information need to be processed in a limited amount of time. Summarisation systems facilitate the extracting of the most relevant information, significantly reducing the time needed to analyse large volumes of text, such as documentation seized in criminal investigations. Summarisation systems can also improve data classification and accessibility, especially in cases where processing by humans would take too long and where precision is not decisive. Legal research We turn now to another use case for NLP technologies: their use in legal research to facilitate the identification of case-relevant statutes, provisions and case-law. While this might be dispensable for research on the law of the EU Member States or non-EU countries posting Liaison Prosecutors to Eurojust (with Eurojust here fulfilling its role as a knowledge hub), there are instances where knowledge of foreign law is necessary for Eurojust to make informed decisions concerning data protection. We refer to the assessments of appropriate safeguards, provided for in Article 56 of the Eurojust Regulation. Knowledge about foreign data protection regulations applicable in the transfer of operational personal data to non-EU countries is an important element of Eurojust’s assessment of existing data protection safeguards. This is a potential use case where AI technology could directly support the application of data protection provisions. Moreover, legal research supported by AI would not require the AI tool to process individuals’ personal data. However, linguistic barriers might be a particular challenge in these situations, making this another case where automated translation could come in handy. The AI Act and Eurojust’s cooperation with third countries Since we have already mentioned Eurojust transfers to third countries, allow us another digression on this point. Some of the solutions proposed by the AI Act might appear complicated when it comes to Eurojust’s relations with external partners. The EC proposes to limit the scope of the AI Act with regard to international law enforcement and judicial cooperation. This would mean that the provisions of the draft AI Act, according to its Article 2(4), would not apply to public authorities in a third country or to international organisations, if these authorities or organisations use AI systems in the framework of international agreements for law enforcement and judicial cooperation with the EU, or with one or more EU Member State. In our view, Article 2(4) would not, in any way, limit the application of the AI Act to EUIs; only public authorities in third countries and relevant international organisations could ‘benefit’ from this proposed exception. The practical application of such an exception in the Eurojust environment raises some questions. While the AI Act would be applicable to Eurojust as it develops or uses AI systems, does this exemption mean that it would not (formally) be applicable to third countries’ Liaison Prosecutors operating at Eurojust? We feel this issue merits some further reflection in advance of the negotiations between legislators. Prior assessment and data protection by design and by default While the EDPS takes note of new and emerging ideas, it is not our intention, nor our role, to plead for these ideas to be put in place. Prior to the set-up of these technologies, authorities considering such applications should perform a legal and ethical assessment to take into account the impact and any possible risk to the fundamental rights and freedoms of individuals, as well as their ethical and legal implications. It is also important to conduct the testing and evaluation of these technologies to ensure that their performance meet the relevant standards, especially regarding data accuracy and bias. The processing of personal data is often at the heart of AI technologies. At the same time, the data collected, processed and stored in judicial systems may be highly sensitive, revealing intimate details about individuals or even causing a threat to their lives. Giving access to this data for the purpose of training algorithms has to be considered with extreme caution and under very strict conditions. Training, testing and validation of machine-learning models with operational personal data and for their further use in the context of a specific Eurojust activity should not be carried out before a data protection impact assessment is done, according to Article 89 of the EUDPR. In addition, we stress that the responsibility of the controller goes beyond that: it starts with adequate project governance, which should take into account the principle of data protection by design throughout the conception and development of the AI tool and system in question. A data-protection compliant AI tool or system can be achieved once the following are in place: clear commitments to this principle in the key documents of the project; policies, processes and methodologies that consider data protection at each stage of the project; by identifying privacy and data protection stakeholders; by assigning roles and responsibilities regarding data protection; by working with competent individuals; and by properly documenting all of these steps. Furthermore, sets of business-level requirements on data protection and mechanisms to assess compliance of the outcome are needed. The controller also needs to put in place procedures for the identification and elimination of any bias in the data used to further train AI models, and to verify that the training data used does not cause discrimination. Processes to check the training or validation of data sets must be built and documented, and procedures allowing for regular monitoring of the models regarding biases and their readjustment or retraining must exist. These processes should include statistical checks on the input and output data. Final remarks From an EDPS perspective, we can clearly see the added value of AI. AI solutions can help complete tasks in a much faster and more cost-effective way, and can also be more accurate and precise than humans, if deployed correctly. At a time when nearly all judicial systems are facing a backlog of cases to process, the promises of efficiency that AI brings cannot be ignored. AI can also detect duplicated information in a reliable way, which contributes to data minimisation and helps to reduce personal data processing by effective anonymisation. If correctly put in place, AI may help to reach true equality and improve access to impartial and objective justice. Nevertheless, we also see the associated risks. Algorithms are only as good as their programmers and the data they have been trained on. This leaves AI systems vulnerable to human error or historical bias. Gains in speed and efficiency can easily turn into disadvantages, if personal data is collected and processed in an immanently biased way. Lack of human oversight and monitoring mechanisms may have dire consequences for the fundamental rights of individuals, as well as their trust in judicial systems and in the EU mechanisms supporting them. Finally, we see many actors in the field trying to be the first to seize the potential benefits of AI. There is a need for a coordinated approach at EU level when it comes to EUIs’ development and use of AI systems to support law enforcement and judicial cooperation. You can count on the EDPS to play its part in the EU’s coordinated approach to AI. Wojciech Wiewiórowski European Data Protection SupervisorMichał Fila Legal Officer at the European Data Protection Supervisor [1] COM(2020) 710 final; Communication of 2 December 2020 from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions ‘Digitalisation of justice in the European Union – a toolbox of opportunities’. [2] COM(2021) 206 final; Proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (Artificial Intelligence Act) and amending certain union legislative acts. [3] Article 2 and 3 of the draft AI Act. [4] The proposed AI Act would explicitly qualify ‘AI systems intended to assist a judicial authority in researching and interpreting facts and the law and in applying the law to a concrete set of facts’ as a high-risk system subject to a particular legal regime (see Annex III, point 8 (a)). [5] EDPB-EDPS Joint Opinion 5/2021 on the proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (Artificial Intelligence Act), 18 June 2021 (available on the EDPS’s website). [6] Regulation (EU) 2018/1725 of the European Parliament and of the Council of 23 October 2018 on the protection of natural persons with regard to the processing of personal data by the Union institutions, bodies, offices and agencies and on the free movement of such data, and repealing Regulation (EC) No 45/2001 and Decision No 1247/2002/EC, OJ L 295, 21.11.2018, pp. 39–98. [7] In particular, Chapter IV of the Eurojust Regulation. [8] Directive (EU) 2016/680 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data by competent authorities for the purposes of the prevention, investigation, detection or prosecution of criminal offences or the execution of criminal penalties, and on the free movement of such data, and repealing Council Framework Decision 2008/977/JHA, OJ L 119, 4.5.2016, pp. 89–131. [9] Article 2(2) and 2(3) of the Regulation 2018/1725. [10] Recital 29 of the Eurojust Regulation. [11] With some notable exceptions, such as the list of prohibited AI practices in Article 5 of the AI Act. [12] For more details on data protection by design as an enforceable legal obligation, see the EDPS Preliminary Opinion no. 5/2018 on Privacy by Design (available on the EDPS’s website). [13] Small Data’s Big AI Potential (available on the Center for Security and Emerging Technology’s website). [14] See also the EDPS Toolkit on Assessing the necessity of measures that limit the fundamental right to the protection of personal data, as well as the EDPS quick guide to necessity and proportionality (available on the EDPS’s website) [15] Cross-border Digital Criminal Justice, Final Report (pp. 160-162) (available on the Publications Office of the European Union’s website). [16] COM(2021) 756 final; Analytical supporting document accompanying the proposal for the Regulation on the JIT collaboration platform (available on the European Commission’s website) [17] See for instance the Connecting Europe Facility Digital programme (CEF Digital) with the eTranslation tool (available on the European Commission’s website).
📘 Today's #Eurojust20 essay:
'Eurojust’s contribution to the fight against terrorism'
- Ilkka Salmi, EU Counter-Terrorism Coordinator
Introduction The judicial dimension plays a key role in countering terrorism, and Eurojust is instrumental in facilitating cooperation in this context. My first meeting in office on 1 October 2021 was with the President of Eurojust, Ladislav Hamran. Since then I have visited Eurojust and participated in its annual counter-terrorism meeting. This has reinforced my belief in the added value of Eurojust. I commend the Eurojust’s solid response to evolving terrorist phenomena based on the rule of law and its commitment to protecting citizens in the European Union and beyond. As the EU Counter-Terrorism Coordinator, I look ahead most of the time. Today I am looking back at how Eurojust has improved the judicial dimension of the fight against terrorism over the years. In paying tribute to those who have striven to achieve significant progress, one can only feel optimistic about future endeavours. My legal and judicial background make me well aware of the impact of the criminal chain on both the repression and prevention of serious forms of criminality. I know how important it is for practitioners to be well-equipped for investigations, prosecutions, convictions, the enforcement of sentences and the reintegration of former convicts. Given the international nature of terrorism and the opportunities provided by the internet and terrorist travel, there is a clear need for international judicial cooperation. Terrorist networks operate transnationally. Incitement to terror, preparatory acts and attacks often take place on the territory of more than one state. Perpetrators and victims often have different nationalities. As a result, many, if not most, terrorism cases require the competent authorities of different states to collaborate. Tackling terrorist phenomena in an effective way would not have been possible without a radical departure from traditional mutual legal assistance. The system of judicial cooperation in criminal matters established within the European Union over the past 25 years is unique in the world. At its heart lies the principle of mutual recognition of judicial decisions. Eurojust, which celebrates its 20th anniversary this year, is one of its most effective tools; it offers Member States and partner countries a modern platform for judicial cooperation. I will first outline the development of Eurojust and the contribution it has made in terrorism cases. I will then turn to the future by highlighting priorities for Eurojust related to counter-terrorism. I will conclude by stressing Eurojust’s role in implementing the priorities I set out at the beginning of my current mandate. The development of Eurojust and its contribution to countering terrorism The Tampere European Council of October 1999 set out the vision of a judicial cooperation instrument that could provide operational support in cases of serious cross-border organised crime. The shock of the attacks on 11 September 2001 contributed to the creation of Eurojust on 28 February 2002[2], followed four months later by the Framework Decision on the European arrest warrant[3] (allowing the surrender of Salah Abdeslam[4] from Belgium to France in two months whereas it had taken ten years for Rachid Ramda[5] to be surrendered from the United Kingdom to France). Over the years, the cooperation of national judicial authorities via Eurojust has contributed to building mutual trust, and the EU judicial cooperation toolbox has expanded to include other mutual recognition instruments, including the Directive on the European Investigation Order[6], which was due to be transposed in 2017. Eurojust’s legal framework has been strengthened[7], and so have its capacities. The Agency has become a key player in facilitating cooperation between the national judicial authorities of the Member States, but also with third countries and other partners. It has supported bilateral and multilateral cooperation. With regard to counter-terrorism, Eurojust’s role has undergone a sea change since 2015. The assistance Eurojust provided to the investigation of the Paris terrorist attacks of 13 November 2015 was pivotal, with 15 Member States involved, as well as the United States; 17 coordination meetings were held in multiple formats and a joint investigation team (JIT) set up including France, Belgium, Eurojust, Europol and later the Netherlands. The experience reinforced the credibility of the Agency as a unique platform for facilitating operational cooperation when many States are involved. While Eurojust’s assistance was requested for a total of only 51 terrorism cases in 2014, this number has progressively increased over the years: Eurojust supported a total of 74 terrorism cases in 2015, 124 in 2016, 178 in 2017, 191 in 2018, 223 in 2019, 217 in 2020 and 221 in 2021[8]. These figures include investigations into major terrorist attacks in EU Member States (such as the Charlie Hebdo and Paris attacks, the attacks in Brussels in March 2016, and the attack on the Christmas Market in Berlin in December 2016) or in third countries (such as the attacks against the Bardo Museum in Tunis in March 2015, and in Ouagadougou in January 2016), but also cases of financing of terrorism, recruitment and training with a view to committing terrorist acts, participation in or support for terrorist groups, production and dissemination of terrorist propaganda, as well as travel to or return from a conflict zone. The cases involved both networks and individuals including high-profile targets. The transmission of requests for mutual legal assistance and for mutual recognition instruments and their execution through Eurojust has proven to be beneficial, particularly in urgent cases. The Agency has supported national judicial authorities facing various challenges relating to, inter alia, the gathering and admissibility of evidence, e-evidence and financial investigations. Eurojust’s assistance and coordination mechanisms have played an essential role, allowing for seizures, confiscations, arrests and convictions in complex cross-border investigations and prosecutions. They have also facilitated the protection and support of victims of terrorism, requests for assistance to third countries as well as the settlement of jurisdictional issues. In 2014, only four coordination meetings were organised in terrorism-related investigations. This number has expanded since then with 15 coordination meetings being held in 2015, 18 in 2016, 14 in 2017, 20 in 2018, 24 in 2019, 12 in 2020 and 9 in 2021. These coordination meetings have provided incomparable added value. In adapting their formats to operational needs, they have brought together magistrates and investigators, allowing them to share in real time and in their own language all useful information with their colleagues, and to define investigation strategies collectively while avoiding duplication or jeopardising parallel initiatives. Additionally, the creation of two operational coordination centres has made it possible to hold successful joint action days: one in 2015 and one in 2017. Eurojust also gave organisational and financial support to a total of 2 JITs in terrorism cases in 2014, 5 in 2015, 6 in 2016, 13 in 2017, 12 in 2018, 8 in 2019, 7 in 2020 and 9 in 2021[9]. JITs, such as the one set up in January 2022 by Sweden and France with the support of Eurojust for proceedings involving core international crimes committed by foreign terrorist fighters against the Yezidi population in Syria and Iraq, are key tools for Member States, but also increasingly for third countries, allowing them to share information and exchange evidence in an efficient manner without the need for a European Investigation Order or mutual legal assistance request, as well as to coordinate investigative measures and prosecution strategies. Information exchange is crucial in terrorism cases, and has been stepped up considerably in recent years regarding both ongoing criminal investigations and prosecutions and proceedings that have already been concluded. The Counter-Terrorism Register, launched in September 2019 on the initiative of the Ministers of Justice of France, Germany, Spain, Belgium, Italy, Luxembourg and the Netherlands, and based on Council Decision 2005/671/JHA on the exchange of information and cooperation concerning terrorist offences[10], has contributed to this in a significant manner. The Register has allowed for the identification of links between prosecutions, and for the detection of the need for multilateral coordination in a number of cases, even when operational cooperation was not facilitated by Eurojust. Eurojust’s efforts aimed at consolidating the uniform and consistent transmission of information, the timely processing thereof, efficient follow-up and regular updates are to be commended. Eurojust has also fostered its collaboration with Europol, in particular with the European Counter Terrorism Centre created in January 2016. The feedback provided by Eurojust to national authorities through, inter alia, the Terrorism Convictions Monitor and ad hoc analyses of landmark court decisions in terrorism cases has also facilitated prosecutions by putting forward comparative legislation, comparative case law and lessons learnt. Additionally, Eurojust has provided strategic input to the Council and its preparatory bodies, on topics such as foreign terrorist fighters, e-evidence and encryption. With contributions to Europol’s TE-SAT report, for example, Eurojust has made it possible to map and analyse trends. Challenges in the gathering of, timely access to and admissibility of battlefield evidence have limited the number of convictions for terrorist offences and international crimes. These challenges have oriented prosecution strategies towards indictments for participation in terrorist organisations, even in cases where this qualification does not guarantee the full accountability of perpetrators and adequate justice for victims. The work of the Eurojust Genocide Network has been crucial in enhancing the use of battlefield information in prosecutions and in encouraging cumulative prosecutions for international crimes and terrorism offences. The cooperation among Member States’ and Eurojust’s affiliated prosecutors on war crimes, as well as the close relations with international partners and NGOs, is internationally referenced as exemplary, inter alia in the context of the Global Counter-Terrorism Forum or within the United Nations. The 2020 Eurojust Memorandum on Battlefield Evidence showed a recent increase in the number of cases based on battlefield evidence and cumulative charges. This is a positive development in the fight against impunity. The Eurojust Memorandum was highly successful in flagging difficulties and disseminating best practices. An additional contribution by Eurojust to strengthening investigations into and prosecutions of terrorism cases is its cooperation with third countries, such as the United States, Norway, Switzerland, Western Balkan states, Turkey and Ukraine. International cooperation has been reinforced through international agreements, Liaison Prosecutors and Contact Points. Around one quarter of the terrorism investigations and prosecutions assisted by Eurojust in 2019 involved non-Member States. Way ahead: counter-terrorism priorities for Eurojust Information sharing is key in countering terrorism. I very much welcome the progress Eurojust has achieved in that area with the Counter-Terrorism Register. It is crucial that Member States share in a systematic and timely manner information on all terrorism-related investigations and convictions, with regular updates to facilitate the establishment of links in proceedings with potential cross-border implications. The Counter-Terrorism Register has already demonstrated in practice that it can strengthen coordination and speed up actions against suspects. Optimising its efficiency depends on the systematic entry and updating of information by the national judicial authorities of all Member States. Further progress is needed. I therefore fully support the proposal made by the Commission in its ‘Security and justice in the digital world’ package[11], published in December 2021, to strengthen digital information exchange on cross-border terrorism cases. It would be beneficial to modernise the Eurojust Case Management System while integrating the Counter-Terrorism Register and its functionalities (especially the link identification), and to set up secure digital communication channels between the competent authorities and Eurojust. Eurojust’s contribution to the digitalisation of justice, in a more general sense, is to be commended. I believe for instance that the creation of a Joint Investigation Teams Collaboration Platform, as proposed by the Commission in the same digital package of December 2021, would bring added value in terrorism cases, by inter alia facilitating the daily management of teams, as well as the collaboration with third countriesand other partners. It would also ensure the secure exchange of information and evidence, the traceability of which would be reinforced. New technologies are key for investigations and prosecutions. I hope that Eurojust will actively participate in the EU innovation hub for internal security at Europol to identify security threats related to new technologies, assess the impact of new technologies on prosecutions and develop innovative tools in joint projects to maximise the use of new technologies in the judicial dimension. The input from magistrates on legal and practical challenges is very important in guiding policymakers. It is also important that Eurojust continues to develop its collaboration with partners such as the European Counter Terrorism Centre at Europol and third countries. The fact that the Council authorised the Commission to negotiate cooperation agreements with 13 more states is also an encouraging sign. Furthermore, Eurojust’s participation in the EuroMed Justice Programme on promoting criminal justice cooperation between the EU Member States and Southern Mediterranean countries is very positive. The full implementation of Directive (EU) 2017/541 on combating terrorism is fundamental. In that context, enhancing judicial authorities’ timely access to information from conflict zones is critical. Eurojust and the Genocide Network should continue their excellent and unique work with the national correspondents for terrorism and international crimes on battlefield evidence and cumulative prosecutions. We are on the right track and I encourage the national competent authorities to build on this to make further progress. Eurojust’s role in implementing the EU Counter-Terrorism Coordinator’s priorities On taking up my duties in October 2021, I set out four priorities for my office, which will evolve over time depending on developments. My first priority is the implementation of the Afghanistan CT Action Plan[12], designed with the Member States, the European Commission, the European External Action Service, the relevant Justice and Home Affairs agencies and international partners, and welcomed by the Council in October 2021. The terrorist threat to the EU is not likely to increase immediately but may grow in the medium term. We need to be prepared and mobilise the existing instruments. There is a role for Eurojust to play as far as the prosecution of foreign terrorist fighters, battlefield evidence and tackling organised crime are concerned, including in cooperation with partner countries such as those in the Western Balkans. My second priority is enhanced assistance to camps and prisons in north-east Syria, where former Da’esh fighters and their families are held. The EU does not intervene in repatriation, to which the Member States take different approaches. Aid and the prevention of further radicalisation in the camps and prisons must be ensured for international and EU security, with a particular focus on minors. The work on battlefield information is important in this context too: building the capacities of national authorities to fight against impunity by obtaining and using battlefield information in court would expand Member States’ options for tackling the challenging legacy of Da’esh. Since the EU is also working towards decongesting the camps by supporting reintegration in local communities in Syria and Iraq, capacity building of national authorities in the region is also critical. The comparative experiences, lessons learnt and best practices issued by Eurojust provide valuable material in this context. Prevention of radicalisation is my third priority. Our work includes projects with a special focus on young people. Investing in education, culture, sports and international exchanges as elements of social cohesion is extremely important. Addressing the ideologies behind violent movements is also necessary: we must look into the roots of Islamist, right-wing and left-wing terrorism and violent extremism from all angles. Terrorism motivated by Islamist extremism remains the main threat we are facing in the EU, but the threat of right-wing violent extremism and terrorism is on the rise. Violent right-wing extremists are increasingly interconnected in the international online space, which exacerbates the threat they pose. I therefore attach great value to the work carried out by Eurojust on this phenomenon. The online spread of terrorist speech, hate speech and disinformation is particularly concerning. The Regulation on addressing the dissemination of terrorist content online[13] and the proposed Digital Services Act[14] are major steps towards addressing this. Major digital companies can and should do much more to curb this phenomenon. Not only should they invest more resources in removing illegal content and moderating harmful content, but they should also refrain from increasing the visibility of divisive and polarising content. I am particularly concerned about the algorithms used by companies such as Facebook and Google to amplify extreme or sensationalist content at the expense of moderate, nuanced and mainstream voices. Commercial gain should not come at the price of creating societal vulnerability and jeopardising security. The EU has the opportunity to set an ambitious standard to protect its citizens and I hope that it will build up the necessary means to achieve its ambitions. In the fight against radicalisation online, cooperation between Eurojust and Europol in the framework of the Scientific Information Retrieval Integrated Utilisation System (SIRIUS) project is very important. The SIRIUS project guides prosecutors to the relevant point of contact for online platforms in the course of their investigations, and sensitises online platforms to the need to build up resources and streamline processes to better respond to judicial requests. Sharing information and best practices on mutual legal assistance procedures and internet-based investigations is beneficial for the national judicial authorities’ capacity building and for online service providers’ outreach. My fourth priority relates to new and disruptive technologies. It is twofold. On the one hand, we should restrict the malicious use of such technologies. While recent terrorist attacks in Europe have so far been low-tech, there is a risk that terrorists will attempt to use new technologies, such as drones, 3D printing and large-scale cyber operations, in future attacks. On the other hand, we should make sure that our law enforcement and security services are equipped with advanced technologies to fight terrorism in full respect of our fundamental freedoms. Eurojust’s contribution in mapping the evolution of criminal practices and the impact of technological changes on prosecution, for instance through its collaboration with Europol in the Observatory Function on Encryption, is important. It makes it possible to have foresight. Further initiatives of this nature should be undertaken. Like my predecessor Gilles de Kerchove, I believe that security and judicial practitioners in Brussels do not have a strong enough voice. There is a risk that technical and legal capabilities to collect information and use evidence in terrorism cases will be severely affected by restrictions on data retention and the use of artificial intelligence, by the increase in end-to-end encryption of electronic communications, further expanded by the roll-out of 5G, and by challenges related to e-evidence. I am keen to contribute to the debate to make sure that we strike the right balance between privacy and security. Here again, Eurojust’s engagement is important, as it echoes legal and practical challenges related to the judicial dimension of the fight against terrorism. Conclusion Eurojust plays a crucial role in supporting the major prosecutions of terrorism in the EU. It provides a strong and modern platform for bilateral and multilateral judicial cooperation, and for information-sharing in counter-terrorism. Like my predecessor, I look forward to working closely with Eurojust. I will strongly support the Agency and I am committed to ensuring that it has the necessary legal framework and adequate resources to provide optimal support to national judicial authorities. Ilkka Salmi EU Counter-Terrorism Coordinator [1] [1] The opinions expressed in this article are those of the author alone and do not necessarily reflect the positions of the Council of the European Union or the European Council. [2] Council Decision 2002/187/JHA of 28 February 2002 setting up Eurojust with a view to reinforcing the fight against serious crime. [3] Council Framework Decision 2002/584/JHA of 13 June 2002 on the European arrest warrant and the surrender procedures between Member States. [4] Facing trial in France for his alleged participation in the terrorist attacks in Paris on 13 November 2015. [5] Convicted in France for the terrorist attacks on public transport in Paris in the summer of 1995. [6] Directive 2014/41/EU of the European Parliament and of the Council of 3 April 2014 regarding the European Investigation Order in criminal matters. [7] The Regulation of the European Parliament and of the Council on the European Union Agency for Criminal Justice Cooperation, replacing and repealing Council Decision 2002/187/JHA was adopted on 6 November 2018 and became applicable on 12 December 2019. [8] Number of new terrorism cases supported by Eurojust per year: 14 in 2014, 41 in 2015, 71 in 2016, 92 in 2017, 84 in 2018, 95 in 2019, 69 in 2020 and 80 in 2021. [9] Number of new JITs supported by Eurojust in terrorism cases per year: 1 in 2014, 3 in 2015, 2 in 2016, 9 in 2017, 2 in 2019, 2 in 2020 and 4 in 2021. [10] Council Decision 2005/671/JHA of 20 September 2005 on the exchange of information and cooperation concerning terrorist offences. [11] Modernising judicial cooperation (09/03/2022) (available on the European Commission’s website). [12] ST 11556/1/21 REV 1, 29 September 2021. [13] Regulation (EU) 2021/784 of the European Parliament and of the Council of 29 April 2021 on addressing the dissemination of terrorist content online. [14] Proposal for a Regulation of the European Parliament and of the Council on a Single Market For Digital Services (Digital Services Act) and amending Directive 2000/31/EC.
📘 Today's #Eurojust20 essay:
'About the Gavanazov II and HP judgments of the CJEU on the European Investigation Order Directive: strengthening the judicial protection in the issuing Member State'
- Anne Weyembergh, Professor, @ULBruxelles
Introduction Like Eurojust, initially set up by the Council Decision of 28 February 2002[1], the framework decision on the European Arrest Warrant and surrender procedure[2] celebrates its 20th anniversary this year. The latter has been the subject of many judgments of the Court of Justice of the European Union (hereafter CJEU). Some of these are very striking ones, such as those relating to the extent of mutual trust and the control to be exercised by the executing authorities, to the possibility of refusing execution on the basis of the risks of infringement of fundamental rights[3], or to the notions of issuing and executing judicial authorities. Some of these judgments have had a significant impact on the functioning of judicial cooperation in criminal matters and on national laws. This is for instance the case of the CJEU judgment in the OG-PI case[4]. Although adopted later, Directive 2014/41/EU of 3 April 2014 on the European Investigation Order[5] is one of the main instruments of mutual recognition in criminal matters in the European Union. Despite its more recent entry into force, this directive has given rise to an increasing number of judgments by the CJEU. Some of them are striking as well. This is particularly the case for the two judgments which will be at the heart of this contribution in honour of the 20th anniversary of Eurojust. These are, on the one hand, the judgment of 11 November 2021, in case C-852/19, Gavanozov II[6] and, on the other hand, the judgment of 16 December 2021, in case C-724/19, HP[7]. The provisions of the Directive which these two judgments interpret are different. The first is essentially linked to Article 14 related to legal remedies, while the second mainly relates to its Article 2 c) i) concerning the designation of the issuing authority. However, both preliminary rulings present common features. Besides the fact that they have both been issued on a referral by the Specialised Criminal Court of Bulgaria (Spetsializiran nakazatelen sad), they have also in common the importance of their impact on the functioning of judicial cooperation and on the national laws of the Member States. Both should result in a strengthening of the judicial protection in the issuing state. I will start with the Gavanozov II case (1) and continue with the HP one (2). The Gavanozov II judgment or the consecration of the right to an effective remedy and of the principle of effective judicial protection The Ivan Gavanazov case concerned criminal investigations into large-scale VAT fraud. The Bulgarian authorities wished to request searches and seizures and a witness hearing by videoconference in the Czech Republic on the basis of a European Investigation Order (EIO). However, under Bulgarian law, there is neither a legal remedy against the lawfulness of searches and seizures and witness hearings nor against the issuance of an EIO dealing with such investigative measures. In such context, the Specialised Criminal Court of Bulgaria referred preliminary questions to the CJEU, the main one seeking to find out whether the national laws of Member States must provide for the possibility of an appeal against the issuance of an EIO to carry out searches and seizures and to organise the hearing of a witness by videoconference. This gave the Court the opportunity to rule on the scope of Article 14 of the Directive on the EIO read in conjunction with the Charter of Fundamental Rights of the EU. As a reminder, Article 14 § 1 of the Directive especially provides that Member States shall ensure that legal remedies equivalent to those available in a similar domestic case, are applicable to the investigative measures indicated in the EIO. In a first judgment dated 24 October 2019 (Gavanozov I)[8] – contrary to Advocate General Yves Bot, who answered all referred questions[9] – the Court reformulated these, considering that the Bulgarian referring Court was simply seeking to know how to complete section J of the form annexed to the Directive. It specified that Article 5, § 1 of the Directive, read in conjunction with the aforementioned section J, must be interpreted as meaning that the judicial authority issuing an EIO must not include in this section a description of the legal remedies available in its national law against the issuance of such an order. That authority needs only to indicate whether a legal remedy has been exercised against the EIO and provide the name and contact details of the competent authorities able to provide further information in this regard. The ‘Guidelines on how to complete the forms’ were then amended accordingly[10]. Being unsatisfied with that first decision by the Court of Justice, the Bulgarian referring court came back with two preliminary questions: a first one intended to know whether national legislation, which does not provide for any legal remedy against the issuing of an EIO for the search of residential and business premises, the seizure of certain items and the hearing of a witness, is compatible with Article 14 of the Directive read in conjunction with Article 47 of the EU Charter of Human Rights, and a second question seeking to discover whether an EIO can be issued under such circumstances. This time, by a judgment of 11 November 2021, the CJEU responded to both questions, generally speaking along similar lines as the former conclusions of Advocate General Yves Bot in Gavanazov I and of Advocate General Michal Bobek in Gavanozov II[11]. In its response to the first question, the Court considered that Article 14 of Directive 2014/41, read in conjunction with its Article 24 (7) and with Article 47 of the Charter, must be interpreted as meaning that it opposes the rules of a Member State issuing an EIO which does not provide for any remedy against the issuance of an EIO having as its object the carrying out of searches and seizures as well as the organisation of a witness hearing by videoconference. In this respect, the judgment highlights the divergences that exist in terms of the level of judicial protection in the various Member States and the lack of harmonisation at the investigative stage. Article 14 of the EIO Directive does not approximate national laws in the field since it limits itself to imposing ‘equivalence’ between legal remedies in domestic cases and investigative measures indicated in EIOs. As the Court stresses: Article 14 ‘does not require Member States to provide additional legal remedies to those that exist in a similar domestic case’ (see point 26). In other words, Bulgarian legislation does not infringe Article 14 of the Directive as such as it does not allow for any remedies against the national investigative measures either. That said, as the Court underlines it, ‘it should be borne in mind that when the Member States implement EU law, they are required to ensure compliance with the right to an effective remedy enshrined in the first paragraph of Article 47 of the Charter, a provision which constitutes a reaffirmation of the principle of effective judicial protection’ (see point 28). Thus, in a way, such a transposition gives the opportunity to the CJEU to transplant by analogy in the context of an EIO the reasoning of the European Court of Human Rights which found repeatedly the absence in Bulgarian law of a legal remedy to domestic investigative measures in breach of the minimum standards under Article 13 of the European Convention on Human Rights[12]. The CJEU refers to these decisions of the Court of Human Rights explicitly (see point 34). In its answer to the second question, the Court considered that Article 6 of Directive 2014/41, read in conjunction with Article 47 of the Charter and Article 4, § 3 TEU, must be interpreted in the sense that it opposes the issuance, by the competent authority of a Member State, of an EIO having as its object the carrying out of searches and seizures as well as the organisation of a witness hearing by videoconference, where the regulations of that Member State do not provide for any remedy against the issuance of such EIO. The Court bases its reasoning on the concept of mutual recognition and mutual trust. As a rule, the executing authority is required to recognise an EIO transmitted in accordance with Directive 2014/41, without any further formality being required, and ensure its execution in the same way and under the same modalities as if the investigative measure concerned had been ordered by an authority of the executing Member State (see points 38 and 39). The mechanism is based on mutual trust and on the rebuttable presumption of compliance by other Member States with Union law and, in particular, with fundamental rights. Observance of those rights falls, primarily, within the responsibility of the issuing Member State, which must be presumed to be complying with Union law and, in particular, with the fundamental rights conferred by that law (see points 54 and 55). Since the absence of legal remedies in the issuing State against the issuance of an EIO infringes Article 47(1) of the Charter, it rules out the possibility of mutual recognition being implemented and benefiting that Member State (see point 56). If an EIO is issued, it would result in the automatic application of Article 11(1)(f) of the Directive which provides for a ground for refusal when there are substantial grounds to believe that the execution of the investigative measure indicated in the EIO would be incompatible with the executing State's obligations in accordance with Article 6 TEU and the Charter (see point 59). In this judgment, the Court thus pays particular attention to the right to an effective remedy and to the principle of effective judicial protection in the issuing state. It strengthens the position of individual rights, which is to be welcomed. The limited control by the executing authorities that results from the mutual recognition principle is to be legitimised by the extent of judicial control on the issuance of the order in the issuing state. As with its case-law on the EAW, faced with sensitive questions[13], the Court paused by initially reformulating the questions put to it but then embarked on a more daring, impacting case-law in terms of the protection of fundamental rights. Among the other important lessons to be learned from this decision is the complementarity between mutual recognition and trust, on the one hand, and the respect for minimum standards and approximation, on the other. This has been underlined by many in the case of the EAW, but it is true also in the field of cooperation at the investigation level, where approximation has remained much neglected by the EU legislator so far. Nothing comparable to the provisions related to the EAW in the Directives on procedural guarantees for suspects and accused persons[14] exists for the moment in the field. Indeed to overcome the divergent approaches of the Member States in the field, the latter has attempted to circumvent differences by leaving a wide margin of discretion to the Member States and extensively referring to national law, including in a number of (nevertheless crucial) aspects of defence rights in transnational investigations such as legal remedies[15]. As was expected by several actors in the field, this judgment is likely to have important consequences. For Bulgaria, of course, it means that as long as Bulgarian law is not made compatible with the Charter and the European Convention of human rights, it will no longer be able to issue EIOs anymore. If Bulgarian EIOs are issued in the same legal context, the executing authorities should apply Article 11(1)(f) of the Directive which provides for an optional ground for refusal based on fundamental rights. The other Member States which do not have such legal remedies in their national law should be impacted as well. It should lead them to revise and correct their national laws accordingly. Logically, the impact should not only concern the legal remedies against the issuance of EIOs but the investigative measures in domestic cases as well. Hence this case-law could have, progressively and indirectly, a positive approximating impact on the level of judicial protection at the investigative stage in the Member States. The decision by the Court leaves many questions unanswered. I will tentatively and without claiming to be exhaustive mention four of them. A first question is to know when the legal remedy should be available, namely before or after the execution of the requested acts and measures (ex ante or ex post). If it is before, the surprise effect will obviously be lost[16]. In other words, if a legal remedy should be available ex ante this would most probably render the issuing of an EIO useless, since the person against whom the investigative measure will be applied will be able to anticipate it. A second question is to know for which other investigative measures such a legal remedy is required. The main criterion for the Court seems to be ‘when a person can be adversely affected’ (see point 47). Of course, this could be interpreted more or less extensively. The Court seems to interpret it rather broadly as it considers that a request to hear a witness by videoconference is also covered by these terms. A third question relates to the precise outlines of the requested reaction and degree of control by the executing authority. It would be particularly detrimental to mutual trust if the executing authority were expected to raise the issue of legal remedy systematically and ex officio and arguably slow down the cooperation in the process, thus making it less effective. The circumstances in which the executing authority should check whether there is any effective remedy available in the issuing State against an EIO should therefore be clarified. A fourth question concerns the exact meaning of the aforementioned statement by the Court according to which the issuance of an EIO in the absence of a legal remedy being available would result in the automatic application of Article 11(1)(f) of the Directive (see supra point 59). The latter only provides for an optional ground for refusal. Does ‘automaticity’, in the view of the Court, imply an obligation to refuse the recognition and execution? At first glance, these two notions (automaticity and optional ground for refusal) would seem to be incompatible. The HP Judgment or the importance of the equivalence principle, the simplification In the case at hand, the Bulgarian public prosecutor’s office had issued four EIOs with a view to collecting traffic and location data associated with telecommunications. Those EIOs were addressed to the Belgian, German, Austrian and Swedish authorities. All orders stated that HP was suspected of financing terrorist activities and that, in the context of that activity, he had had phone conversations with persons residing in the territory of these four Member States. The competent German, Austrian and Swedish authorities did not transmit a decision recognising the EIOs, but the Belgian investigating judge did. On the basis of the evidence gathered, HP was charged, together with other persons, with illegally financing terrorist activities and participating in a criminal organisation seeking to finance those activities. The referring court, once again the Specialised Criminal Court of Bulgaria, before which HP’s indictment was brought, wanted to determine whether that accusation was well founded. Indeed, according to Bulgarian law, these EIOs had been issued by the Public Prosecutor’s Office whereas, in a similar domestic case, the authority with competence to order that traffic and location data associated with telecommunications is a judge of the Court of First Instance, having jurisdiction in the case concerned, whereas the public prosecutor only has the power to make a reasoned request to that judge in such a situation. Hence, it referred two questions to the CJEU. The first question was to find out whether Article 2 (c) (i), of the Directive precludes a public prosecutor from being competent to issue, during the preliminary phase of criminal proceedings, an EIO aimed at obtaining traffic data and location data relating to telecommunications, where, in the context of a similar national procedure, the adoption of an investigative measure aimed at accessing such data falls within the exclusive competence of the judge. The Court answered in the affirmative. It started by analysing the letter of that provision but concluded that it does not allow the Court to respond to the question. The Court thus examined the context and objectives of the said provision (see points 30 and 31). In terms of context, it considers that, in order to assess the necessity and proportionality of an investigative measure – which is a requirement according to Article 6(1) a) of the Directive – and to provide the additional explanations referred to in Articles 26(5), 27(4) and 28(3) of the Directive, the issuing authority must be the investigating authority in the criminal proceedings concerned, which is thus competent to order the gathering of evidence in accordance with national law (see points 32 to 34). Article 6(1)(b) of the Directive which provides that the issuing authority may only issue an EIO where the investigative measure(s) referred to therein could have been ordered under the same conditions in a similar domestic case, leads the Court to consider that only an authority which is competent to order such an investigative measure under the national law of the issuing State may be competent to issue an EIO (see point 35). Turning to the objectives of the Directive, the Court of Justice particularly insisted on its simplification purpose and concluded that a distinction between the authority which issues the EIO and the authority which is competent to order investigative measures in the context of those criminal proceedings would risk complicating the system of cooperation, thereby jeopardising the establishment of a simplified and effective system (see points 36 to 38). The Court then replied negatively to the second question, which was to discover whether the recognition of such an EIO by the competent authority of the executing state (public prosecutor or an investigating judge) replaces the court order required under the law of the issuing state. In other words, it considers that the executing authority cannot, by its decision of recognition, remedy the non-compliance with the conditions for issuing an EIO (see point 50). The opposite solution would indeed affect the distribution of competences between the issuing authority and the executing authority, and thereby the balance of the EIO mechanism based on mutual trust, since this would amount to recognising the executing authority with the power to control the substantive conditions for issuing such a decision (see points 51 to 53). In this decision, besides putting the emphasis on the equivalence principle requiring the application of the same rules as in a similar domestic case, the Court insisted on the simplification purpose that lies at the heart of the EIO directive as well as on the balance of roles between issuing and executing authorities and the restricted control only that can be performed by the latter. In this case, the result is in a way favourable to judicial protection since it indeed results in a judge issuing EIOs. As highlighted by the Court itself, it is also in line with another judgment by the Court dated 2 March 2021, in case Prokuratuur C‑746/18, related to Directive 2002/58/EC of 12 July 2002 concerning the processing of personal data and the protection of privacy in the electronic communications sector. In this decision, the Court considered indeed that Article 15(1) of that Directive read in the light of Articles 7, 8 and 11 and Article 52(1) of the Charter of Fundamental Rights, must be interpreted as precluding national legislation that confers upon the public prosecutor’s office, whose task is to direct the criminal pretrial procedure and to bring, where appropriate, the public prosecution in subsequent proceedings, the power to authorise access of a public authority to traffic and location data for the purposes of a criminal investigation. In its HP judgment, the CJEU repeats the part of Advocate General’s conclusions where he noted that an EIO seeking to obtain traffic and location data associated with telecommunications cannot be issued by a public prosecutor where that public prosecutor not only directs the criminal pretrial procedure but also is in charge of the public prosecution in subsequent criminal proceedings (see points 42 and 43). Conclusion These two cases show the major added value that CJEU case-law can bring in interpreting aspects of EU legislation in this field and especially in terms of the need to find a balance between security and efficacy, on the one hand, and protection of human rights and judicial protection, on the other. The added value of the functioning of the EAW was already clear. This is now true of the EIO Directive as well. These two judgments show replies to some important questions, but they are far from exhaustive. Others will surely follow, which will allow the Court to bring further necessary clarifications. In a way, this case-law is also representative of Eurojust’s added value. Its impact on operational cooperation is of course well known. As stated in Eurojust’s 2021 Annual Report, the Agency dealt with 4 262 cases involving an EIO in 2021, and helped to resolve issues concerning challenges with the execution of EIOs for the hearing of suspects or accused persons via videoconference, or the interception of telecommunication[17]. Its role in terms of informing and disseminating important decisions of the CJEU and analysing them, and the impact these decisions have on the functioning of judicial cooperation and on national legislations, is perhaps less known. Yet the same Annual Report mentions that ‘Eurojust also monitors relevant CJEU case-law developments in the field of the EIO directive and their possible impact on judicial cooperation’. This is crucial in the EIO context but also with respect to the other mechanisms of judicial cooperation in criminal matters[18]. Together with the European Judicial Network, Eurojust has indeed a key function to fulfil. Not only do they facilitate the implementation of EU legislation in the field of judicial cooperation in criminal matters as such, but equally they assist national authorities by monitoring CJEU judgments, drawing attention to them, helping authorities to understand these judgments and adjust their practice to the implications of these judgments, and identifying best solutions and practices. This is essential for practitioners – including defence lawyers – and academics involved and interested in the implementation of EU legislation in this field. Anne Weyembergh Professor, Université Libre de Bruxelles [1] Council Decision of 28 February 2002 setting up Eurojust with a view to reinforcing the fight against serious crime, OJ L 63, 6.3.2002, pp. 1-13. [2] Framework Decision 2002/584/JHA of 13 June 2002 on the European arrest warrant and the surrender procedures between Member States, OJ L 190, 18.7.2002, pp. 1–20. [3] CJEU, 5 April 2016, C-404/15 and C-659/15 PPU, Aranyosi and Caldararu, ECLI:EU:C:2016:198; CJEU, 25 July 2018, C-220/18 PPU, ML; CJEU, 25 July 2018, C-216/18 PPU, LM, ECLI:EU:C:2018:586; CJEU, 15 October 2019, C-128/18, Dumitru-Tudor Dorobantu, ECLI:EU:C:2019:857 and CJEU, 17 December 2020, C-354/20 PPU and C-412/20 PPU, L&P, ECLI:EU:C:2020:1033. [4] CJEU, 27 May 2019, OG - PI, C-508/18 and C-82/19 PPU, ECLI:EU:C:2019:456. [5] Directive 2014/41/EU of 3 April 2014 regarding the European Investigation Order in criminal matters, OJ L 130, 1.5.2014, pp. 1–36. [6] ECLI:EU:C:2021:902. [7] ECLI:EU:C:2021:1020. [8] C-324/17, ECLI:EU:C:2019:892. [9] See Opinion of Advocate General Y. Bot, 11 April 2019 (ECLI:EU:C:2019:312). [10] See Council of the EU, doc. 5291, 23 January 2020. [11] See Opinion of Advocate General Y. Bot, points 51-90 and Opinion of Advocate General, 29 April 2021 (ECLI:EU:C:2021:346). [12] ECtHR, 22 May 2008, Iliya Stefanov v. Bulgaria, CE:ECHR:2008:0522JUD006575501, § 59; ECtHR, 31 March 2016, Stoyanov and Others v. Bulgaria, CE:ECHR:2016:0331JUD005538810, §§ 152 to 154; and ECtHR, 19 January 2017, Posevini v. Bulgaria, CE:ECHR:2017:0119JUD006363814, §§ 84 to 86). [13] See CJEU, 29 January 2013, C-396/11, ECLI:EU:C:2013:39, where the Court also reformulated the questions to then embark in more daring judgments (see its Aranyosi and Caldararu case-law as referred to in fn 3). [14] See Directive of 20 October 2010 on the right to interpretation and translation in criminal proceedings; Directive of 22 May 2012 on the right to information in criminal proceedings (letter of rights); Directive of 22 Oct. 2013 on the right of access to a lawyer in criminal proceedings and on the right to communicate upon arrest (‘Salduz Directive’); Directive of 11 May 2016 on procedural safeguards for children who are suspects or accused persons in criminal proceedings; Directive of 26 October 2016 on legal aid for suspects and accused persons in criminal proceedings and for requested persons in European arrest warrant proceedings. [15] In this regard, see E. Sellier and A. Weyembergh (eds), Criminal Procedures and Cross-border Cooperation in the EU Area of Criminal Justice. Together but Apart? Ed. De l’Université de Bruxelles, 2020, p. 312 and fn. [16] In this respect, see especially Opinion of Advocate General Michal Bobek, points 56 and 57. [17] Eurojust 2021 Annual Report, p. 24. See also Report on Eurojust’s casework in the field of the European Investigation Order. [18] See especially Eurojust, Case-law by the Court of Justice of the European Union on the European Arrest Warrant.
Eurojust
EU Fundamental Rights ➡️ #Huma
EDPS