Daniel Evanko
banner
evanko.bsky.social
Daniel Evanko
@evanko.bsky.social
Director of Journal Operations at the American Association for Cancer Research (AACR). Opinions my own.
The new COPE guidelines on retraction that lists “undisclosed involvement of AI” as a form of misrepresentation that could cause a loss of confidence in the work that justifies retraction will almost certainly act as an incentive to disclose.
September 25, 2025 at 10:44 PM
Certainly from some authors.
September 25, 2025 at 1:29 PM
I don’t think the requirements will change much in the near future but I expect some efforts to confirm the validity of the AI-use disclosures.
September 25, 2025 at 11:30 AM
Pangram does retrain regularly. It seems that usually there is little change between model releases, but sometimes there is a substantial change in a model’s output that makes the retraining critically important.
September 12, 2025 at 2:47 AM
Yes but writing is not the rate limiting step, completing a publishable research study is. The writing can be completed in under a week w/o AI. The study takes months and often years. So speeding up the writing will not increase the load on reviewers as long as journals filter out the junk first.
July 24, 2025 at 12:26 PM
Interesting, but your premise appears to ignore the time required to actually do the research, what the underlying cause(s) of the recent increase in submissions is, or whether ethical use of AI could aid editorial/peer review more than it aids creation of legitimate research outputs.
July 24, 2025 at 10:37 AM
Looking forward to presenting this work at #PRC10. Use of genAI by authors and reviewers is increasing rapidly and the ability to measure it with high sensitivity and a false positive rate <0.1% is invaluable for tracking usage.
July 16, 2025 at 7:37 PM
You are correct that the current policies at virtually all journals regarding immediate public access of the author manuscript through green OA don’t satisfy upcoming US funder requirements. The embargo period will need to be eliminated, but I doubt any will do so until required.
March 13, 2025 at 9:18 PM
None of the public access plans released by US funding agencies in response to the Nelson memo require gold open access, only immediate public access. So this requirement was not expected to negatively impact societies in a substantial way.
March 13, 2025 at 11:32 AM