Google staff resign in backlash over its AI contracts with military

A number of engineers and researchers at Google DeepMind have resigned or are leaving in protest over the company’s recent military-focused AI contracts, citing ethical and safety concerns. Named departures include Sofia Liguori and Andreas Kirsch, who say Google’s expanding work with US and Israeli defence organisations - including a Pentagon agreement that permits use of its models for classified purposes and previous involvement in Project Nimbus - conflicts with the company’s stated AI principles on human oversight and human rights. Google says it follows consistent review processes, acceptable-use policies and internal channels for raising concerns, but departing staff argue those safeguards are insufficient and that contracts limit employees’ ability to speak openly without risking retaliation.

Critics highlight practical risks of deploying current AI in warfare: models are seen as not robust enough, governance and visibility into how the Pentagon uses the technology are claimed to be weak, and AI can compress decision time, undermining meaningful human review in high-pressure settings. The debate follows broader moves by other major tech firms to provide AI to defence projects and references past systems such as Project Maven and Palantir integrations, underscoring tensions between rapid military adoption, ethical guardrails, and the potential reshaping of tech workforce morale and recruitment.