Google researchers demand freedom to tell the world what they know about AI
Some Google DeepMind staff in Britain say employment terms restrict public criticism after they leave the company.
TL;DR
- Google DeepMind employees in Britain are asking the company to remove non-disparagement terms that restrict public comments about AI risks.
- The Washington Post reviewed a contract and interviewed two current workers; Google said it does not enforce the clauses after employees leave and has removed that restriction for new hires.
- Current employees want changes to existing contracts, and the terms’ enforceability and the company’s response to those requests remain open.
The Washington Post reported that some employees in Google’s London-based AI division are asking executives to release them from non-disparagement terms so they can speak publicly about AI risks. The story was also discussed in a Hacker News thread linking to the Post report; that discussion documents reader reaction, not independent confirmation of the contracts. [1][2] [1] [2]
The Post said it reviewed one employment contract and interviewed two current Google DeepMind workers. They said some British employees’ agreements prohibit criticizing the company even after leaving; the employees spoke anonymously because they feared retaliation. [1] [1]
Google told the Post that it would not enforce the post-termination restriction against former employees and that it had updated terms for new hires to remove it. The workers quoted by the Post said they want the company to remove the clause from existing contracts as well. [1] [1]
The report leaves two questions unresolved: whether the existing restrictions would be legally enforceable and whether Google will revise current employees’ agreements. The Post noted that British law protects qualifying whistleblowers who raise matters in the public interest, but the workers said the clause could still deter them from speaking. [1] [1]
Why it matters
The dispute concerns how employment terms shape public scrutiny of AI development. Google’s stated policy addresses former employees and new hires, while the Post reports that current workers want clarity and changes to existing contracts.
Editor's note
The Washington Post is the reporting source; Hacker News is cited only as a scored discussion of that linked report, not as independent evidence for the contract claims.