More evidence of the corrosive impact of AI on #scholarship #integrity. According to this research in the biomedical context, #fabricated #citations existed in: ~1 in 2,828 papers (2023), ~1 in 458 (2025), ~1 in 277 (early 2026). A twelvefold increase.
Horrifying. The recommendations here are obvious (e.g. #retractions); but I might also suggest that sanctions against authors who are found to have submitted such work are necessary (e.g. desk rejection).
Update. Alan Colin-Arce has written an excellent slide deck on multilingual publishing in the humanities and social sciences.
https://sisu.ut.ee/wp-content/uploads/sites/903/Humanities-and-SS-Commons.pdf
#OpenPharma just published a vision statement calling on #pharma companies to make their research #OpenAccess.
https://bmjopen.bmj.com/content/16/5/e104128
The vision statement calls for OpenAccess to texts, #OpenData (under #FAIR principles), open #PIDS (#ORCID and #ROR), plain language summaries, and patient involvement.
It understands that the goal is a stretch. and distinguishes easier short-term goals from harder long-term goals, while arguing for both.
OpenPharma is funded by voluntary contributions from pharma companies, and considers itself a collaboration of "a number of important stakeholder groups: the pharmaceutical industry, publishers, patients, academics, regulators, editors, non-pharmaceutical funders and societies."
https://www.openpharma.blog/wp-content/uploads/2021/02/2020-01-15-Charter-of-Open-Pharma.pdf
Also see the OpenPharma home page.
https://www.openpharma.blog/
Open Science aims to fight misinformation and improve trust in scientific research; it encourages the reliability and accessibility of evidence, reduces inequalities through the democratisation of scientific knowledge and focuses scientific endeavours on issues of societal significance. As a multisponsor collaboration committed to driving positive change, Open Pharma has a multifaceted vision for scientific research publications funded by pharmaceutical companies (‘company research publications’) that aligns strongly with Open Science tenets. This new vision statement outlines our forward-looking principles for company research publications, both for short-term attainment (‘immediate’) and long-term commitment (‘ultimate’). Together, the principles provide a framework for positive collective action by all stakeholders involved in the development and dissemination of peer-reviewed company research publications. Underpinned by our central commitment to transparency for company research publications, we outline goals for: universal access to these publications; provision of peer-reviewed plain language summaries of the publications to aid comprehension among non-specialist readers; leveraging author and institutional metadata to advance transparency, discoverability and research impact; working towards FAIR (Findability, Accessibility, Interoperability, and Reuse) data principles through cross-sector consensus and action, and disclosure of patient involvement in research and its reporting to support transparency and encourage a research ecosystem attuned to patient centricity. We call on all stakeholders to realise the Open Pharma Vision and achieve an open and trusted future for company research publications that will ultimately advance patient care and improve global health.
As referências fabricadas se multiplicaram na literatura biomédica de 2024 para cá. Os geradores de lero-lero estão obviamente envolvidos, mas no meu entendimento eles estão principalmente evidenciando más práticas que já existiam antes.
https://blog.leonardof.med.br/2026/ciencia-fabricada.html
Curiosidade: o levantamento que identificou as referências fabricadas usou um gerador de lero-lero em uma etapa da análise.
"The _Astrophysical Journal_ accepted our #AI's [fast radio burst] discovery after three rounds of peer review, but the AAS editorial office then halted the paper over an AI-disclosure review, not the science."
https://blankline.org/newsroom/ai-frb-paper-apj-halted
RE: https://mastodon.social/@Daojoan/116530684580599648
“We build legitimate institutions around legitimate problems - but the world changes. The institution remains optimised for the old problem, and gradually becomes ceremonial. Like the British monarchy. Or the United Nations. Or the academic peer review system.”
From the #NSF budget request
$10 million in FY 2027 for "Publication Fee Alternative Exploration"
Are AI overviews quietly breaking our research discovery ecosystem? If zero-click readership becomes the norm, who controls visibility, what does impact even measure, and can we still claim to be “engaging” with research at all?
#AI #scholcomm #research
George Macgregor
petersuber
Leonardo Fontenelle
Peter Suber
ilkayholt
Karen Keiller 🥌 📚
Kate Nyhan
Jens Mittelbach