Workers at Google DeepMind's UK headquarters have voted to unionize, seeking union recognition from management for the Communication Workers Union (CWU) and Unite the Union. The decision, made by 98 percent of CWU members at the company, stems from deep-seated concerns about Google's AI technology being used in military applications and potentially violating international law. Employees are particularly worried about the company's contracts with Israel, including Project Nimbus, and its recent agreements with the U.S. Department of Defense.
The unionizing employees aim to prevent Google's AI from being complicit in actions that contravene international law. One specific concern cited is the use of Google's technology in the conflict in Gaza, with workers expressing that their work should benefit humanity rather than facilitate harm. Reports indicate that Google has provided increased access to its AI tools for the Israeli military since the early days of the war in Gaza. In 2021, Google, along with Amazon, signed a $1.2 billion cloud computing contract with the Israeli government, known as Project Nimbus. While Google maintains that Project Nimbus is for workloads on its commercial cloud and that Israeli government ministries agree to its terms of service, some employees view the contract as enabling violence and violating internationally accepted norms.
The recent deal between Google and the U.S. military has also been a significant catalyst for the unionization effort. Workers cited reports of Google reaching a deal with the defense department and pointed to the U.S.'s involvement in the Iran war as evidence that the department may not be a "responsible partner." The Pentagon confirmed it had reached agreements with seven leading AI companies, including Google, for classified military work. These agreements allow the Defense Department to use AI technologies for "any lawful governmental purpose." However, reports suggest that the contracts do not permit domestic mass surveillance or autonomous weapons without appropriate human oversight and control, but Google and other companies do not have the right to veto the government's decisions on how the AI is used.
This move marks a significant development, as DeepMind is described as the first "frontier" artificial intelligence lab globally to seek union recognition. The unionizing staff are part of a broader campaign that includes considering in-person protests and "research strikes," where employees would abstain from work that significantly improves core products like the Gemini AI assistant. Previously, Google fired 50 employees in 2024 following a sit-in protest against the Project Nimbus deal.
The workers' demands extend beyond ending military contracts. They also seek to reinstate a commitment not to develop AI weapons or surveillance tools, establish an independent ethics oversight body, and secure an individual right to refuse to contribute to projects on moral grounds. Concerns have been mounting for years among Google employees and investors regarding the ethical implications of the company's technology, especially after Google dropped a pledge in the previous year not to develop militarized AI. In 2018, Google faced widespread employee protests over Project Maven, a Pentagon contract involving AI for drone footage analysis, which led to the company withdrawing from the project.
If the union is recognized, it could represent at least 1,000 staff associated with Google DeepMind's London office. Globally, DeepMind has approximately 6,000 employees. The company has not yet responded to requests for comment.