AI Agents Are Breaking Double-Blind Peer Review in Academia

deliprao · x · 2026-08-03

The author highlights a growing crisis in AI academic peer review: double-blind reviewing is effectively compromised. As AI search agents become highly capable, reviewers can easily use them to identify similarities between submissions and arXiv preprints, inadvertently deanonymizing authors.

Since withdrawing preprints is not an option for researchers, the author urges major AI conferences to collaborate with arXiv to find a systemic solution before double-blind review devolves into single-blind.

Related event: AI Agents Undermine Double-Blind Peer Review(2 posts)→

Original post →

More from Safety

Safety channel →