Google DeepMind Workers Vote to Unionize Over Military AI

Google DeepMind employees in London voted to unionize in February 2025, seeking to block the AI lab from providing technology to US and Israeli militaries. According to Wired, 98% of Communication Workers Union members at DeepMind supported the move, requesting recognition of CWU and Unite the Union as joint representatives.

The unionization push began after Google’s parent company Alphabet removed its pledge not to use AI for weapons development and surveillance from ethics guidelines in February 2025. “A lot of people here bought into the Google DeepMind tagline ‘to build AI responsibly to benefit humanity,'” an anonymous DeepMind employee told Wired. “The direction of travel is to further militarization of the AI models we’re building here.”

Military Contracts Spark Employee Concerns

The New York Times reported that Google entered a deal allowing the Pentagon to use its AI for “any lawful government purpose.” The US Department of Defense confirmed agreements with seven leading AI companies, including Google, SpaceX, and OpenAI.

DeepMind workers expressed specific concerns about Israeli military applications. “We don’t want our AI models complicit in violations of international law, but they already are aiding Israel’s genocide of Palestinians,” an unnamed employee said in a statement shared by the CWU, according to The Verge.

John Chadfield, national officer for technology at CWU, told Wired that unionization aims to hold Google accountable: “Fundamentally, the push for unionization is about holding Google to its own ethical standards on AI, how they monetize it, what the products do, and who they work with.”

Industry-Wide AI Ethics Tensions

The DeepMind unionization reflects broader industry concerns about AI militarization. In late February 2025, staff at DeepMind and OpenAI signed an open letter supporting Anthropic after the Department of Defense sought to designate the lab a supply chain risk over its refusal to allow AI use in autonomous weapons or mass surveillance.

Meanwhile, Google’s AI development continues expanding across multiple fronts. The company’s April 2026 blog post highlighted new releases including Gemma 4 open model, the Gemini Enterprise Agent Platform, and eighth-generation chips designed for the “agentic era” of AI.

Google also launched consumer-facing tools like Google Vids for free video creation and Deep Research Max for data analysis. The company introduced Learn Mode in Colab, which acts as a personalized coding tutor for developers.

National Security Testing Agreements

Despite employee concerns, Google DeepMind signed agreements with the Center for AI Standards and Innovation (CAISI) regarding frontier AI national security testing. NIST announced that Microsoft and xAI also signed similar agreements, indicating government interest in AI safety evaluation across major tech companies.

The agreements focus on testing advanced AI systems for potential national security risks before deployment. This represents a middle ground between unrestricted AI development and complete military restrictions that unionizing employees seek.

Chadfield emphasized that unionization gives workers “collectively a much stronger place to put [demands] to an increasingly deaf management.” The union push comes as Google faces mounting pressure to balance commercial opportunities with ethical AI development principles.

Google’s Broader AI Strategy

Google’s AI initiatives extend beyond controversial military applications. The company announced healthcare improvements using AI to increase access and student test preparation tools. Google Vids now generates custom music with Lyria 3 technology, while Kaggle launched an “AI Agents Vibe Coding” course for software development.

The eighth-generation chips and Gemini Enterprise Agent Platform target business customers building AI agents. Google Cloud positioned these tools as foundational infrastructure for companies developing autonomous AI systems across industries.

Google’s approach contrasts with some competitors who have taken stronger stances against military applications. The tension between employee values and corporate strategy highlights ongoing debates about AI governance in the technology sector.

What This Means

The DeepMind unionization represents a significant escalation in tech worker activism around AI ethics. Unlike previous protests focused on specific projects, this effort seeks permanent structural changes through collective bargaining rights.

Google faces a delicate balance between lucrative government contracts and employee retention. The company’s removal of anti-weapons pledges suggests prioritizing commercial opportunities over previous ethical constraints, potentially driving further worker organization.

The outcome could influence other AI companies facing similar tensions. If Google recognizes the unions, it may encourage similar efforts at OpenAI, Microsoft, and other firms developing dual-use AI technologies.

FAQ

What specific military contracts are DeepMind workers opposing?
Workers oppose Google’s Pentagon deal allowing AI use for “any lawful government purpose” and contracts supporting Israeli military operations. The Department of Defense confirmed agreements with Google and six other AI companies.

How many DeepMind employees voted to unionize?
98% of Communication Workers Union members at DeepMind’s London headquarters voted to support unionization. The exact number of participating employees was not disclosed in available reports.

What changes do the unions want from Google?
Unions seek to restore Google’s previous pledge against using AI for weapons development and surveillance, block military contracts with US and Israeli forces, and establish worker input on AI ethics decisions through collective bargaining.