Google DeepMind Employees Clash With Management as Union Talks Stall

Talks between Google DeepMind employees and company leadership ended in a shouting match on Wednesday, not an agreement. The team negotiating on behalf of workers left the meeting frustrated, saying management had no intention of seriously discussing demands related to military contracts and ethical oversight of AI development.

The dispute centers on DeepMind's London office, where nearly 300 employees have signed up with the Communication Workers Union (CWU) and Unite. In a vote among CWU members, 98 percent backed formal recognition of the union by the employer. Organizers ultimately want to bring more than a thousand employees at the lab's UK office into the movement.

What workers are demanding

The union's demands aren't about pay or working conditions in the traditional sense, but about control over what the technology they build is used for. They are calling for a binding ban on weapons and surveillance system contracts, the creation of an independent ethics committee, and the right of individual employees to opt out of projects they consider morally problematic.

The immediate trigger for the dispute was Google's contract with the Pentagon for the use of Gemini models, which allows the technology to be used for military purposes without the ethical restrictions that previously applied. More than 600 Google employees had earlier signed an open letter opposing the deal, and in February 2025 the company dropped a pledge not to use AI in ways that violate international norms on weapons and surveillance.

Wednesday's talks collapse

According to accounts given to the media by meeting participants, DeepMind's leadership did not offer any concrete concessions or a timeline for further talks. One researcher involved in organizing the union said getting a seat at the negotiating table is the only way to regain real influence over the company's decisions. Union organizers said that despite the rocky start, they would keep pushing for dialogue, treating the matter as unresolved.

If successful, the effort would create the first labor union at a leading AI lab anywhere in the world, at a company competing directly with OpenAI and Anthropic for top researchers. That matters because the AI talent market is exceptionally fluid, and engineers and researchers currently hold enormous bargaining power when choosing an employer.

What it means for the industry

The DeepMind dispute fits into a broader trend of employees at major AI labs organizing around questions about the ethical limits of the technology's development, a debate also playing out inside OpenAI and Anthropic. Increasingly, it is not pay but control over military and surveillance applications of the technology that has become the flashpoint between employees and management at tech companies.

For Polish observers of the AI market, this dispute shows that questions about the ethical limits of artificial intelligence, also present in EU debates over the AI Act, concern not only regulators but the people building the technology inside companies. If the London precedent succeeds, it could shape how other AI labs, including those planning to expand in Europe, negotiate employment terms for researchers and engineers.

Further rounds of talks between the union and DeepMind's leadership are expected in the coming weeks. Google has not yet officially commented on Wednesday's meeting or on employees' demands regarding military contracts.

Sources: Google DeepMind Union Talks Hit Wall as Execs Resist (techbuzz.ai), Google DeepMind workers in the U.K. vote to unionize over military AI contracts amid internal backlash over its Pentagon deal (fortune.com).