Oxford AI professor proposes desk-rejecting papers flagged by LLM watermark detection

TuhinChakr · x · 2026-09-23

Yarin Gal suggests conferences desk-reject any paper flagged by LLM watermarking detection, arguing researchers should write their own papers with LLMs used at most for feedback. Bhuwan Dhingra agrees, advocating that only manually written papers be accepted and that raising the submission bar is the right policy even if some find workarounds.

Related event: Researchers propose desk-rejecting papers that fail LLM watermark detection(2 posts)→

Original post →

More from Research

Research channel →