Stanford professor challenges DeepMind biosecurity watermark paper as trivially removable
anshulkundaje · x · 2026-10-04
Professor Anshul Kundaje publicly criticized Pushmeet Kohli's DeepMind biosecurity watermark research. While the paper's discussion acknowledges caveats, he argues evidence from Rohit shows that for the reported results it is trivial to eliminate the watermark — a valid critique he asks the authors to address directly. He finds the algorithm interesting as a pilot study with potential biosecurity applications, but says the paper fails to substantiate its claims.
Related event: Stanford Professor Challenges DeepMind's AI Biosecurity Watermark Paper(4 posts)→
More from Safety
- Minneapolis council votes 7-6 to force drivers in robotaxis; Mayor Frey vetoes it as a backdoor ban — Afinetheorem · 2026-10-04
- OpenAI safety leader quits, warning AI company's culture is 'broken' — jethronethro · 2026-10-04
- Polymarket Opens Bets on Which AI Lab Pauses Training in 2026, OpenAI at 10% — Polymarket · 2026-10-04
- OpenAI Safety Employee Resigns After 3.5 Years, Says Company Culture Is Broken — Polymarket · 2026-10-04
- Reddit essay rebuts Hinton: no testable evidence AI already has subjective experience — WhoReallyKnowsThis · 2026-10-04
- Against Hinton: sounding human and rogue agents aren't evidence of AI consciousness — WhoReallyKnowsThis · 2026-10-04