Google DeepMind staff urge leaders to drop non-disparagement clauses on AI risk warnings
Current DeepMind employees in the United Kingdom have told management that employment contracts should no longer contain non‑disparagement clauses that block former staff from publicly warning about AI safety risks. The staff claim the clauses could silence important warnings even after employees leave the company, potentially hindering broader public awareness of emerging dangers.
Key points
- DeepMind staff in the UK say non‑disparagement clauses block former employees from warning about AI risks.
- Employees urged leaders to remove the clauses to allow public safety disclosures after departure.
- Alphabet has not yet responded to the staff's request, according to the Washington Post.
The request was reported by the Washington Post, citing unnamed workers who said the language in their contracts limits their ability to speak out about AI risks. DeepMind, a subsidiary of Alphabet, has not publicly responded to the allegations. The push reflects growing internal concern that legal restrictions may conflict with broader industry calls for transparency and responsible AI development.
If the company changes its contracts, it could set a precedent for other AI firms facing similar scrutiny over how they handle employee speech on safety issues. The debate also ties into wider regulatory discussions about corporate accountability for AI risk communication.
The story so far
2 episodes →- Google DeepMind staff urge leaders to drop non-disparagement clauses on AI risk warningsthis story
Coverage and discussion
1source- Hacker News discussion · 75 pointsnews.ycombinator.com
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Policy & Regulation
All →- Google to disable AI in K-12 schools unless admins opt in · 10 src
- Anthropic updates Claude to block recreation of existing video games · 30 src
- Anthropic and OpenAI set up election war rooms amid AI misinformation risks · 4 src
- Apple requires dating and AI companion apps to carry 18+ rating · 1 src
- Crypto and AI PACs spent $250 million in 2026 federal elections · 1 src
Comments
via GitHub Discussions