Google experiences resignation backlash over AI military contracts

Google is facing criticism from some employees because the company is using its artificial intelligence (AI) technology for military purposes.

The BBC has spoken to several technology workers who have recently left, or are preparing to leave, Google DeepMind, Google's AI company. They are worried about the company's military contracts.

Sofia Liguori, a research engineer based in London, is one of them. She resigned from Google DeepMind this week and spoke to the BBC about her concerns, even though her contract restricts her from doing so.

"Knowing that my company is pushing more and more to use AI systems in military operations makes me feel ashamed to work at Google," Liguori said. She was speaking for herself, not on behalf of the company.

A Google spokesperson told the BBC: "We work with public sector organisations around the world. We follow the same process when we review and agree to these contracts, including checking that they follow our rules for acceptable use and our AI principles."

"We also have special ways for employees to safely raise concerns, from business ethics to misconduct. We have strict workplace rules to prevent any kind of punishment or unfair treatment in response to these concerns."

Google is one of several technology companies that signed contracts with the US military earlier this year to use AI in secret military settings. US Defence Secretary Pete Hegseth has made adding AI to the armed forces a major goal. He has said the Trump administration wants the US military to become an "AI-first warfighting force".

But AI technology is changing quickly. Critics question whether it is reliable enough to be used in war without more testing, stronger safety rules and human supervision.

Liguori worked at Google for three years before deciding to leave because of its growing work with the US and Israeli militaries. "I don't agree with this direction, and it has made me lose faith that my company is staying true to its stated ethical values," she said.

Google's AI Principles say that people should oversee AI and that the technology should respect "international law and human rights". But current and former employees, along with experts who spoke to the BBC, said they were worried that people might not have enough control over AI at critical moments.

The technology workers said they are especially worried about the AI military contracts signed this year because they greatly expand the ways the Pentagon can use the technology.

Andreas Kirsch, an AI researcher, also left Google DeepMind earlier this month because of his concerns.

Kirsch told the BBC that he believes "Google's rules and checks are not strong or meaningful enough to make sure AI is used safely and responsibly in military and secret work". He also said they were not enough "to develop artificial superintelligence safely in the future".

"Google seems to have no power to control or block how the Pentagon uses its AI, and it also seems to have limited information about how the technology is being used," he added.

Over the past few years, the US military has started using several technologies that include AI to improve its ability to fight wars.

The best-known example is Maven Smart Systems, the Pentagon's main software system for organising intelligence and other data.

US defence company Palantir developed the system. It uses AI to help military staff find and choose targets to attack. Palantir has added technology from several leading AI companies to Maven, allowing some tasks to be completed much faster.

Retired Lieutenant General Jack Shanahan, who used to lead Project Maven, said the approach seemed to be: "Move as fast as possible to use [AI technology] as widely as possible, possibly at the cost of careful and strict testing and evaluation."

However, the Pentagon says using AI is essential to winning wars. "Speed defines victory in the AI era," Emil Michael, an under secretary in the US Department of Defense, said in a statement earlier this year.

Google's contract with the Pentagon, signed earlier this year, allows its AI technology to be used for secret military work. The agreement says Google's AI models can be used for "any lawful government purpose". It also says they "should not be used for domestic mass surveillance or autonomous weapons (including target selection) without appropriate human oversight and control".

OpenAI, Amazon, Microsoft, SpaceX, Oracle, Nvidia and the start-up Reflection signed similar contracts. These agreements allow their AI technology to be used in different military settings, including transport, planning and battlefield operations.

As AI technology develops quickly, more people are raising concerns about its use in war zones.

"AI shortens the time a person has to make a decision because it works much faster than a human. That is especially risky in a high-pressure situation," said Sarah Shoker, who used to lead OpenAI's Geopolitics team.

"When AI chooses hundreds of targets and there is very little time, it is almost impossible for a human operator to properly check the AI's results for mistakes."

Google employees have told the BBC that they are afraid to speak out about AI safety and its use in war. They say their employment contracts prevent them from publicly criticising the company.

"In principle, I'm not allowed to criticise Google for the rest of my life," Sofia Liguori told the BBC. But she decided to speak out "because I believe it's important for people to know about these things, and I also think that talking about facts is not criticism".

Liguori added that at Google, "people are afraid of being punished and putting their careers at risk. You feel that you are being watched. You can't speak."

A current Google DeepMind employee, who did not want to be named because they feared punishment, told the BBC: "Many employees who joined because they believed in the company's original mission may leave. But there are also people at the start of their careers who could take their places. It could be a changing of the guard."

Another Google DeepMind employee, who also wanted to remain anonymous, told the BBC: "Many people who strongly disagree with how Google DeepMind has changed its ethical position have left the company."

"Some people are fighting back. Others are thinking: 'It's a big company. What do you expect to change?'" the employee said.

Google's agreements with Israel have also caused controversy. In 2024, employees criticised the company over Project Nimbus, a $1.2bn (£960m) contract it shares with Amazon. The project provides Israel's government and military with AI technology and cloud computing services. At the time, Google said Project Nimbus "is not directed at highly sensitive, classified, or military workloads relevant to weapons or intelligence services".

Israel's use of AI in warfare has also been debated after a recent documentary called Naza examined how AI-powered systems choose targets for attacks on the Gaza Strip, where many civilians have been killed. An Israeli government minister said the claims made in the film were false and misleading.

Andreas Kirsch, a former Google employee, believes AI technology is not ready to be used on the battlefield. "Today's AI models are not reliable enough, and I don't think they have been trained for these kinds of tasks either," he said.

Rosa Rahimi also contributed reporting.