Google DeepMind Staff Push to Kill Contracts That Reportedly Gag Them for Life
The Washington Post reports some Google DeepMind staff in Britain have contracts that bar them from criticizing the company even after they leave. Google says it will not enforce the clause on former staff, but current employees want it deleted from existing contracts.
Imagine leaving a job and still being contractually barred from saying a bad word about your old employer for the rest of your life. According to The Washington Post, that is what some Google DeepMind staff in Britain signed up for, and a growing number of them want it gone.
The Post says it reviewed a contract and spoke to two current DeepMind workers, who asked for anonymity to avoid retribution. Per the report, non-disparagement clauses in some UK employment contracts keep applying after people leave. The two workers say the terms stop staff from talking to the public about the potential impacts of AI.
The row follows our earlier report on DeepMind researchers quitting over military AI controls.
The spark, the Post reports, was a former DeepMind employee telling the BBC last month that she had chosen to ignore her contract to speak out against Google building military AI. Current UK staff then went looking at their own paperwork. One of the employees told the Post that some of them found the same kind of clause. Dozens then piled into an internal discussion thread. Some criticised the clause directly. Others posted an emoji of the text 1984, a nod to Orwell.
"If it's not enforceable then the reason is to basically intimidate people." One current DeepMind employee, speaking anonymously to The Washington Post
That employee told the Post they do not think the lifetime restriction is enforceable. The Post notes that British law protects whistleblowers who speak out in the public interest. Nobody has reported a tribunal ruling on this clause, and there is no report that Google has ever used it against anyone. So the legal question is untested, and the employee's view is an opinion, not a finding.
That is the problem staff are pointing at. A clause that probably would not survive a court can still do its job in a quiet corridor. A researcher weighing whether to go public with a safety worry may not want to find out the hard way.
Google says it will not enforce it
Google did respond. The Post says a Google executive told the internal thread on Friday that the company would not enforce the clause against workers who had left. In an emailed statement on Tuesday, a spokesperson said:
"We've always encouraged an open dialogue about safety and responsibility. Recently, we've updated our terms for new hires, which included removing the post-termination aspect of non-disparagement clauses." A Google spokesperson, in an email to The Washington Post
The spokesperson added, per the Post, that Google does not enforce such clauses after people leave and has dedicated internal channels for workers to escalate concerns to top leaders.
A verbal promise from an executive and a fixed template for new hires are real moves. But look at who they cover. New hires get cleaner terms. Former employees get a pledge of non-enforcement. The people still on existing contracts are the group the reported changes do not touch. The Post says those workers want the clause dropped from existing contracts entirely, and one source says some want updated language to give them legal certainty. Neither the Post nor Google says those contracts will be amended.
A promise not to enforce something is cheap. Rewriting the document is what counts.
The Post points out that executives and rank and file staff at Anthropic and OpenAI regularly speak about AI danger in public, while fewer voices from Google do. A lifetime gag clause is one plausible reason for that quiet, though nobody in the reporting says it is the reason. What DeepMind does with the contracts of its current London staff will say more than any spokesperson statement.
