Philipp Hacker

@philipphacker@eupolicy.social
111 Followers
37 Following
145 Posts
Prof for Law and Tech at Eur. Uni Viadrina, ENS / Fellow at UCL
passionate about AI and mountains
https://www.europeannewschool.eu/philipp-hacker.html
Working onAI and law; data protection and privacy; algorithmic fairness; explainable AI; platform regulation
also seriousabout food and wine
Can't wait for this conference, Tue & Wed in Geneva. 'Legal and technical challenges of large generative ai models' https://www.europeannewschool.eu/genaiconference23 Organised by @philipphacker & Sarah Hammer (University of Pennsylvania Law School, USA)
#chatgpt @llms #ai #tech #law
genaiconference23 - European New School of Digital Studies

Study and research on the digital transformation | PhD program | Fellowship program ▶ Readyto shape digital Europe? Join ENS!

2/2
Sandra is a true thought leader and one of the world's most prolific and respected writers on the subject, and we are glad and honored to have her with us in Berlin this month.

More information on the talk: https://www.weizenbaum-institut.de/events/sandra-wachter-protecting-algorithmic-groups-under-anti-discrimination-law/

Link to the paper: https://arxiv.org/abs/2205.01166

VIDEO Sandra Wachter - The Theory of Artificial Immutability: Protecting Algorithmic Groups under Anti-Discrimination Law

As part of our Distinguished Fellow Program, Prof. Sandra Wachter of the Oxford Internet Institute presented her work on Discrimination and AI.

Weizenbaum Institut

1/2
Don't miss Oxford's Sandra Wachter and her talk on AI and discrimination this Wednesday (June 21) at 6 PM CEST, jointly hosted by Weizenbaum-Institut für die vernetzte Gesellschaft and European New School of Digital Studies. Join us in person at Jacob-und-Wilhelm-Grimm-Zentrum in Berlin or online! Drinks will only be provided offline, though :)

Registration (video link will be sent): https://survey.lamapoll.de/Registration_SandraWachter/

#ai #aiact #responsibleai #discrimination

Registration_SandraWachter (LamaPoll - Online surveys)

5/n
Ultimately, law, computer science and sustainability perspectives – both in academia and in industry – need to converge to effectively address the dual large-scale transformations of AI and sustainability.

cc: @ens
#aiact #sustainableai #climate #climatechange #ai #ml #ki #it #itlaw

4/n
For example, the subjective right to erasure, in some situations, has to be balanced against the collective interest in mitigating climate change. And if retraining an AI model to comply with the right to be forgotten entails the annual energy costs of a small town, we should perhaps refrain from granting that right – or at least factor these sustainability costs into the equation.

3/n

(2) This regulatory toolkit may then, in a second step, serve as a blueprint for other information technologies and infrastructures facing significant sustainability challenges due to their high GHG emissions, for example: blockchain; Metaverse applications; and data centers.
(3) Most importantly, certain rights we have come to take for granted, such as the right to erasure (Article 17 GDPR), may have to be limited due to sustainability considerations.

2/n
In this, the paper explores two key dimensions: legal instruments to make AI greener; and methods to render AI regulation, and the law itself, more sustainable:

I try to make three concrete points:

(1) First, regulating for more sustainable AI: we need to go beyond mere transparency. I consider co-regulation strategies; sustainability by design; restrictions on training data; and consumption caps. One important strategy going forward may be: sustainability impact assessments.

1/n
Extremely excited that my new paper is out: Sustainable AI Regulation (https://arxiv.org/abs/2306.00292)! And that the paper was discussed at the one and only Privacy Law Scholars Conference, I couldn’t have thought of a better debut...

Given the deplorable state of our planet, the law (including IT law) must adapt, and in certain areas quite radically. Hence, for AI regulation, we need a shift from trustworthiness to sustainability.

Sustainable AI Regulation

Current proposals for AI regulation, in the EU and beyond, aim to spur AI that is trustworthy (e.g., AI Act) and accountable (e.g., AI Liability) What is missing, however, is a robust regulatory discourse and roadmap to make AI, and technology more broadly, environmentally sustainable. This paper aims to take first steps to fill this gap. The ICT sector contributes up to 3.9 percent of global greenhouse gas (GHG) emissions-more than global air travel at 2.5 percent. The carbon footprint and water consumption of AI, especially large-scale generative models like GPT-4, raise significant sustainability concerns. The paper is the first to assess how current and proposed technology regulations, including EU environmental law, the General Data Protection Regulation (GDPR), and the AI Act, could be adjusted to better account for environmental sustainability. The GDPR, for instance, could be interpreted to limit certain individual rights like the right to erasure if these rights significantly conflict with broader sustainability goals. In a second step, the paper suggests a multi-faceted approach to achieve sustainable AI regulation. It advocates for transparency mechanisms, such as disclosing the GHG footprint of AI systems, as laid out in the proposed EU AI Act. However, sustainable AI regulation must go beyond mere transparency. The paper proposes a regulatory toolkit comprising co-regulation, sustainability-by-design principles, restrictions on training data, and consumption caps, including integration into the EU Emissions Trading Scheme. Finally, the paper argues that this regulatory toolkit could serve as a blueprint for regulating other high-emission technologies and infrastructures like blockchain, Metaverse applications, and data centers. The framework aims to cohesively address the crucial dual challenges of our era: digital transformation and climate change mitigation.

arXiv.org

Squaring the regulatory circle: Here is the video of my statements and answers at the public hearing on generative AI, held at the German Parliament. In the exciting hearing, we discussed the state, risks and benefits of generative AI, the AI Act of course, content moderation, and other things of interest.

Link: https://youtu.be/ExvSVPQlQFE

cc: @ens
#gpt4 #aiact #chatgpt #itrecht #ai #ki

Philipp Hacker: Anhörung im Deutschen Bundestag zum Thema „Generative Künstliche Intelligenz“

YouTube

3/3
direct link to the German version of my statement (which was basically machine-translated with DeepL due to lack of time on my end):
https://www.bundestag.de/resource/blob/949660/d3645b4b8c892713bc3b5633d86969a8/Stellungnahme-Hacker-DE-data.pdf

#gpt4 #chatgpt #aiact #dsa #ai #ki #itrecht #itlaw #riskmanagement

cc: @ens