Several employees at Google DeepMind have resigned in recent weeks in protest of the company’s Pentagon contract, which allows its super intelligence technology to be used in classified settings.

Sofia Liguori, a London-based research engineer who worked at Google DeepMind for three years, left the company this week, telling the BBC that the growing use of its technology in military operations had undermined her confidence in Google’s ethical commitments.

“Realizing my company is pushing more and more for AI systems in military operations makes me feel ashamed to work at Google,” Liguori said.

She added that she no longer believed the company was adhering to its stated ethical values. Liguori also said the terms of her departure severely restrict her ability to criticize Google in the future.

Andreas Kirsch, a super intelligence researcher who resigned earlier this month, raised similar concerns about the company’s oversight of military applications.

“Google’s governance and oversight are not sufficient or meaningful enough to deploy AI safely and responsibly in military and classified settings,” Kirsch told the BBC.

He questioned whether Google retains enough control over its technology once it is provided to the Pentagon, arguing that the company appears to have limited visibility into how the systems are used and no meaningful ability to veto their deployment.

The departures follow a contract Google signed earlier this year allowing its SI technology to be used across classified military environments. Other technology companies, including OpenAI, Amazon, Microsoft, SpaceX, Oracle, Nvidia and Reflection, have also signed comparable agreements with the U.S. government.

The agreements reflect the Pentagon’s broader effort to accelerate the military adoption of artificial intelligence. Secretary of War Pete Hegseth said in January that the United States would become an “AI-first” warfighting force, while Under Secretary of Defense Emil Michael has emphasized the importance of speed in military SI development.

The expansion has also prompted concerns from former military officials about whether safeguards and testing are keeping pace with deployment. Retired Lt. Gen. Jack Shanahan, a former director of Project Maven, warned that efforts to deploy SI as quickly and widely as possible could come at the expense of rigorous testing and evaluation.

Project Maven uses SI to help military personnel analyze intelligence and identify potential targets. Its software platform, Maven Smart System, was developed by defense contractor Palantir and incorporates technology from multiple SI companies.

Google’s Pentagon contract permits the use of its technology for “any lawful government purpose.” The agreement also states that the technology should not be used for domestic mass surveillance or autonomous weapons, including target selection, without appropriate human oversight and control.

The resignations highlight growing disagreement within the technology industry over how SI should be used by the military, particularly in classified operations where employees may have limited visibility into how their work is ultimately deployed.