Google DeepMind Employees Resign Over Pentagon AI Deal

Several Google DeepMind employees have resigned in recent weeks over a Pentagon contract signed earlier this year that allows Google AI to be used in classified settings. Sofia Liguori, a London-based research engineer, left the company this week after three years. Speaking to the BBC in a personal capacity despite a contract restriction, she said the company’s growing push to use AI in military operations made her ashamed to work there. Liguori said she had lost faith in Google’s commitment to its stated ethical values. She also said her departure terms mean she is, in principle, barred from criticizing Google for the rest of her life.

Andreas Kirsch, an AI researcher, resigned earlier this month over similar concerns. He told the BBC that Google’s governance and oversight were not sufficient for safe and responsible AI use in military and classified settings. Kirsch said he hopes to develop artificial superintelligence safely, and questioned whether Google can control or veto the Pentagon’s uses of its AI or see how it is deployed.

Google signed the contract with the US military earlier this year to deploy AI across classified environments. OpenAI, Amazon, Microsoft, SpaceX, Oracle, Nvidia and startup Reflection have signed comparable agreements. The Pentagon is accelerating AI adoption; War Secretary Pete Hegseth said in January that the United States would become an “AI-first” warfighting force, while Defense Department under secretary Emil Michael said this year that “speed defines victory in the AI era.”

Retired Lieutenant General Jack Shanahan, a former director of Project Maven, criticized the Pentagon’s goal of deploying AI as widely and quickly as possible, potentially at the expense of rigorous testing and evaluation. Maven Smart Systems, the Pentagon’s main software platform for organizing intelligence and data, was developed by Palantir and uses AI to help personnel identify and select targets. Palantir has incorporated technology from several leading AI companies into Maven. Google’s contract permits AI use for “any lawful government purpose,” but says it should not be used for domestic mass surveillance or autonomous weapons, including target selection, without appropriate human oversight and control.

Breitbart previously reported that Kirsch criticized the agreement in a post on X on Tuesday, calling the signing “shameful.” More than 600 Google employees had also sent CEO Sundar Pichai a letter urging him to block classified military use of the company’s AI. Kirsch told Business Insider he had hoped the letter would prompt reconsideration, but woke to news that Google had signed what he called a worst-case version of the contract. Breitbart’s article also includes editorial commentary about Google employees and promotes reporter Lucas Nolan’s book, Code Red.

Several Google DeepMind employees have resigned in recent weeks over a Pentagon contract signed earlier this year that allows Google AI to be used in classified settings. Sofia Liguori, a London-based research engineer, left the company this week after three years. Speaking to the BBC in a personal capacity despite a contract restriction, she said the company’s growing push to use AI in military operations made her ashamed to work there. Liguori said she had lost faith in Google’s commitment to its stated ethical values. She also said her departure terms mean she is, in principle, barred from criticizing Google for the rest of her life.

Andreas Kirsch, an AI researcher, resigned earlier this month over similar concerns. He told the BBC that Google’s governance and oversight were not sufficient for safe and responsible AI use in military and classified settings. Kirsch said he hopes to develop artificial superintelligence safely, and questioned whether Google can control or veto the Pentagon’s uses of its AI or see how it is deployed.

Google signed the contract with the US military earlier this year to deploy AI across classified environments. OpenAI, Amazon, Microsoft, SpaceX, Oracle, Nvidia and startup Reflection have signed comparable agreements. The Pentagon is accelerating AI adoption; War Secretary Pete Hegseth said in January that the United States would become an “AI-first” warfighting force, while Defense Department under secretary Emil Michael said this year that “speed defines victory in the AI era.”

Retired Lieutenant General Jack Shanahan, a former director of Project Maven, criticized the Pentagon’s goal of deploying AI as widely and quickly as possible, potentially at the expense of rigorous testing and evaluation. Maven Smart Systems, the Pentagon’s main software platform for organizing intelligence and data, was developed by Palantir and uses AI to help personnel identify and select targets. Palantir has incorporated technology from several leading AI companies into Maven. Google’s contract permits AI use for “any lawful government purpose,” but says it should not be used for domestic mass surveillance or autonomous weapons, including target selection, without appropriate human oversight and control.

Breitbart previously reported that Kirsch criticized the agreement in a post on X on Tuesday, calling the signing “shameful.” More than 600 Google employees had also sent CEO Sundar Pichai a letter urging him to block classified military use of the company’s AI. Kirsch told Business Insider he had hoped the letter would prompt reconsideration, but woke to news that Google had signed what he called a worst-case version of the contract. Breitbart’s article also includes editorial commentary about Google employees and promotes reporter Lucas Nolan’s book, Code Red.