Negotiations between Google DeepMind and its London-based employees over the prospect of unionization stumbled this week, with staff voicing frustration at what they described as an unwillingness among senior executives to engage meaningfully with the issue.
The talks, held Wednesday, represent the first formal bargaining session since DeepMind workers voted 98 percent in favor of unionizing in May, seeking joint representation by the Communication Workers Union and Unite the Union on behalf of at least 1,000 employees at the company’s UK headquarters. Google management initially declined to recognize the unions but offered to enter negotiations instead, a concession that now appears to have stalled.
According to Wired’s Joel Khalili, who covered the May vote and the subsequent negotiations, employees left Wednesday’s session frustrated. The core demand from workers remains unchanged: a clear commitment from DeepMind leadership not to pursue contracts involving weapons, surveillance, or technology that could cause large-scale harm. Staff are also seeking the right to abstain from projects that violate their personal moral or ethical standards, and a formal process for negotiating the effects of AI deployment on roles and job security.
The unionization push was triggered by DeepMind’s expanding work with military clients, including contracts with the Israeli government. In the vote announcement, organizers stated: “We don’t want our AI models complicit in violations of international law, but they already are aiding Israel’s genocide of Palestinians. Even if our work is only used for administrative purposes, as leadership has repeatedly told us, it is still helping make genocide cheaper, faster, and more efficient.”
The May vote was itself the culmination of more than two years of growing tension at DeepMind over military AI work, during which staff issued open letters, held internal protests, and threatened “research strikes,” abstaining from work on Gemini and other Google AI products as a pressure tactic.
DeepMind is not alone among frontier AI labs in facing organized labor pushback. OpenAI and Anthropic have both seen employees raise concerns about the direction of their work, though DeepMind’s UK location gives its workers unusual legal leverage under British labor law, which provides a more structured path to union recognition than is available in most US states.
The rocky start to negotiations suggests the path to a formal union agreement at DeepMind will be longer and more contentious than some employees had hoped. Under UK law, if management continues to resist voluntary recognition, the unions can petition the Central Arbitration Committee to force a legally binding recognition ballot.
Sources: Wired (Joel Khalili, Jul 3, 2026); The Verge (May 5, 2026)