Two researchers have resigned from Google DeepMind over Google's expanding AI contracts with the US military. Sofia Liguori, a London-based research engineer who joined the lab in 2023, left this week, and AI researcher Andreas Kirsch resigned earlier in October. Both have spoken publicly about why they left, and two other current or departing Google DeepMind staff members described similar concerns to the BBC on condition of anonymity.
The departures center on the AI agreements Google signed with the Pentagon earlier in 2026. Google's contract allows its AI to be used for classified military work, and states that the models can be used for "any lawful government purpose." The classified deal was announced in April. Google is not alone: OpenAI, Amazon, Microsoft, SpaceX, Oracle, Nvidia and Reflection have signed similar agreements with the Defense Department.
Liguori said she objects to Google's growing work with both the US and Israeli militaries, and to the use of AI in targeting and warfare generally, not only to autonomous weapons. She pointed to a non-classified GenAI.mil contract that she tied to Google Cloud's rise to IL6 security authorization, the level used for classified workloads. She also cited Project Nimbus, the $1.2 billion contract Google shares with Amazon to provide AI and cloud services to Israel's government and military. Google has said that contract is not directed at highly sensitive or military workloads.
“I no longer believe that our technology will have a net positive impact on society.”— Sofia Liguori, Former Google DeepMind research engineer
Key facts
- 01Google DeepMind research engineer Sofia Liguori, who joined in 2023, resigned this week over Google's military AI work.
- 02AI researcher Andreas Kirsch left Google DeepMind earlier in October, questioning oversight of Pentagon use of Google's AI.
- 03Google's 2026 Pentagon contract allows its models to be used for "any lawful government purpose," including classified work.
- 04OpenAI, Amazon, Microsoft, SpaceX, Oracle, Nvidia and Reflection have signed similar Pentagon AI deals.
- 05Google says it follows a consistent review process for public-sector contracts and has policies against retaliation.
Liguori's objections predate her resignation. The day after Google announced its classified Pentagon contract in April, she staged a personal sick-out, using sick leave and holiday time. She said the company's internal handling of dissent, including deleted posts and responses from HR, also contributed to her decision to leave. She spoke to the BBC despite an employment contract that she said restricts her from criticizing Google.
Kirsch's concern is narrower and more procedural. He questioned whether Google's governance and oversight are adequate once its models are in military hands, and whether current AI systems are reliable enough for that use. "Current AI models are not robust enough," he said.
“Google seems to have no right to control or veto how the Pentagon uses its AI”— Andreas Kirsch, Former Google DeepMind AI researcher
Google responded by pointing to its contract review process. A Google spokesperson, Tom Hewett, said the company works "with public sector organisations worldwide" and that it follows "a consistent process for reviewing and entering these contracts," including "compliance with our acceptable use policies and AI principles." The company added that it has "dedicated channels for employees to safely raise any concerns" and "strict workplace policies in place to prevent retaliation of any kind."
The Pentagon has been explicit about the value it places on moving fast with commercial AI. Emil Michael, an under secretary at the US Department of Defense, has put it in a single line: "Speed defines victory in the AI era." Sarah Shoker, formerly head of OpenAI's Geopolitics team, has raised the other side of that tradeoff, pointing to scenarios with hundreds of AI-selected targets and a short window to review them.
The two anonymous staff members gave differing reads on what happens next inside Google DeepMind. One said many motivated employees could leave over the issue. Another said some staff are pushing for change internally, while others doubt the company's direction can be altered. Employees also said their contracts discourage public criticism of the company.
Liguori said her decision came down to direction rather than any single contract. "I don't agree with this direction," she said, adding that Google's push toward AI systems in military operations made her "lose faith in my company being true to its stated ethical values."
Liguori's resignation letter urged her former colleagues to "stay human" in the face of military work. Kirsch, for his part, focused his critique on governance rather than on the contracts themselves, arguing that Google lacks the ability to control or veto how the Pentagon uses its models.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
