Google DeepMind Workers Vote to Unionize After Classified Pentagon AI Deal Overrides Eight Years of Ethics Pledges
In 2018, 4,000 Google employees killed a Pentagon contract (Project Maven) that used the company’s AI to analyze drone surveillance footage. In 2026, Google signed a bigger, classified AI deal with the Pentagon for "any lawful governmental purpose." Now, the AI researchers at DeepMind are unionizing.
Key Takeaways:
- Unionization Vote: Workers at Google DeepMind's UK offices voted 98% in favor of joining the Communication Workers Union and Unite the Union in April.
- Demands: The union demands include ending military AI use, restoring Google’s scrapped weapons pledge, and creating an independent ethics body.
- Contrast: In 2018, Google published AI principles pledging not to develop weapons or surveillance technology; eight years later, these principles were removed, and the company fired leaders of its AI ethics team.
In 2026, despite over 580 Google employees signing a letter urging CEO Sundar Pichai to refuse the classified military AI deal, and hundreds demanding DeepMind research not be used for weapons, Google proceeded with the agreement. The union demands are specific:
- End Military AI Use: Stop using Google AI by the Israeli and US militaries.
- Restore Weapons Pledge: Reinstate the company’s commitment not to build AI weapons or surveillance tools.
- Independent Ethics Oversight: Create an independent ethics oversight body.
- Moral Refusal Right: Individual researchers' right to refuse projects on moral grounds.
These demands highlight a disconnect between stated ethical principles and revenue-driven decisions, prompting the need for collective action from DeepMind researchers.