Google DeepMind Workers Vote to Union on Military AI Deals

Google employees Deep Mind In London, they voted to unionize as part of an effort to prevent the AI ​​lab from supplying its technology to the United States and Israel Armies.

In a letter Addressed to Google’s UK and Ireland managing director, Debbie Weinstein, the workers asked the company to recognize the Communications Workers’ Union and Unite the Union as joint representatives of DeepMind employees.

“Fundamentally, the push for unions is about holding Google to its own ethical standards for AI, how it monetizes it, what the products do, and who they work with,” John Chadfield, CWU’s national chief technology officer, tells WIRED. “Through the process of unionization, workers are collectively in a much stronger place to place[demands]on an increasingly tone-deaf management.”

The campaign to unionize began in February 2025, when Alphabet, Google’s parent company Remove pledge Not using artificial intelligence for purposes such as weapons development and surveillance is part of its ethical guidelines, according to a DeepMind employee, who requested anonymity for fear of retaliation.

“A lot of people here have bought into Google DeepMind’s mantra to ‘build AI responsibly for the benefit of humanity,’” one DeepMind employee told WIRED. “The direction of travel is to further militarize the AI ​​models we build here.”

This concern is increasingly reflected throughout the industry. In late February, DeepMind and OpenAI employees I signed an open letter In support of the Anthropists, after the US Department of Defense sought to do so Designate the laboratory as a supply chain risk Because of its refusal to allow its artificial intelligence to be used for autonomous weapons or for mass surveillance of American citizens. Last week, the New York Times I mentioned Google has struck a deal that allows the Pentagon to use its artificial intelligence “for any legitimate government purpose.” (Friday, US Department of Defense certain It has reached deals with seven leading AI companies — including Google, SpaceX, OpenAI, and Microsoft — to use its models on underground networks.) Nearly 600 Google employees are based in the U.S. It is said He signed a letter of protest against the deal. “We think the ‘any legal purpose’ clause is vague enough to be effectively meaningless,” the DeepMind employee says.

Google did not immediately respond to a request for comment. The company has previously defended its deals with government organizations. “We are proud to be part of a broad consortium of leading AI labs and technology and cloud companies providing AI services and infrastructure in support of national security,” said Jane Kreider, a Google spokeswoman. New York Times Last week. “We are committed to the private and public sector consensus that AI should not be used for domestic mass surveillance or autonomous weapons without appropriate human oversight.”

In 2021, Google employees in the United States Formed the Alphabet Workers Union. Alphabet does not recognize the union for collective bargaining purposes, but has done so previously He succeeded in negotiating agreements On behalf of Google’s contractors.

A DeepMind employee told WIRED that if employees succeed in unionizing in the UK, they will likely demand that Google withdraw from its alliance. Long-term contract with the Israeli armyseeking greater transparency about how its AI products are used, and the kind of safeguards surrounding layoffs made possible by automation.

If Google doesn’t participate, employees will ask to do so, the letter states Jury To force the company to recognize unions.

Since the beginning of the year, both Anthropic and OpenAI It announced large-scale expansions of its operations in London. CWU hopes that unionization efforts at DeepMind will motivate workers at those labs to take similar action.

“These conversations happen,” Chadfield says. “Workers at other frontier laboratories have seen what the people at Google DeepMind have done. They have come to us for help, too.”