Google AI staff quit over military contracts, citing ethical and oversight worries

A series of resignations has struck Google DeepMind as engineers such as Sofia Liguori and Andreas Kirsch quit over the firm’s deepening AI collaborations with the United States and Israeli militaries. Liguori, a London-based research engineer, said the push to embed AI in combat operations made her feel ashamed to work for Google, while Kirsch warned that the company’s governance is inadequate for safe military use. Google maintains that it reviews public-sector contracts against its AI Principles and provides channels for staff to raise ethical issues without retaliation.

The contracts, signed earlier this year, permit Google’s models to be used for any lawful government purpose, though critics highlight the Pentagon’s fast-track AI strategy, exemplified by Project Maven and Project Nimbus, as risking insufficient testing and human oversight. Experts cited concerns that AI-driven targeting could compress decision times beyond what operators can safely manage. Internal sources say many employees feel constrained from speaking out, fearing career repercussions.