Findings of ACL already separates papers from talks, so 'AI-authored' rules are moot, scholar argues

ipeirotis · x · 2026-09-29

In a discussion with @johnjhorton, ipeirotis argues venues already have mechanisms akin to 'AI papers not presented to humans': ACL's Findings track accepts papers that are correct but not deemed worth presenting, and other conferences have similar policies. The real problem, he says, is that 'AI-authored' is an ambiguous notion that will be abused in both directions — and in any case, he predicts such distinctions will become irrelevant within a year or two.

Original post →

More from AGI Musings

AGI Musings channel →