Google DeepMind: Ethical Tensions and Military Pressures

A Departure Motivated by Ethics

In January, tragic events involving agents of the Department of Homeland Security (DHS) caused the deaths of two people. In each case, a federal agent used their weapon against an unarmed citizen, with fatal consequences.

Faced with these incidents, Alexander Matt Turner, then a researcher at Google DeepMind, discovered that Google was providing Cloud services to DHS agencies, which deeply shocked him. He considered it unacceptable that federal agents could act this way, and sought to influence his company to stop working with these agencies. His initiative quickly widened into a campaign to keep Google from entering into unethical military agreements, as the Pentagon increased pressure on AI providers for unrestricted collaborations on controversial technologies.

His goal was to ensure that Google DeepMind’s (GDM) ethical commitments would hold up under pressure. He especially wanted GDM to maintain its position against supporting killer robots. He asked many influential figures, known for their commitment to AI ethics and safety, for help. However, almost all of them declined to get involved.

Take Stuart Russell, a renowned AI researcher who has been campaigning against autonomous weapons for more than ten years. The researcher worked in his lab for several years. At a conference, Russell publicly agreed to push his organization to support AI providers against government pressure and promised to consult its members. But neither the statement nor the survey ever materialized.

Another example is Jeff Dean, Google’s Chief Scientist and co-lead of the Gemini AI project. In 2018, Jeff had signed a commitment never to support the development or use of killer robots. The researcher persuaded him to co-sign an amicus brief in support of Anthropic against the Pentagon. However, he also asked him to use his influence to keep Google from entering into an unethical military agreement, but he doubts that he did. Dean stayed at Google despite his commitment.

The researcher drew up a 25-page proposal, including contractual clauses and oversight mechanisms. Experts in military law and oversight praised the proposal, which represented an ethical alternative that Google could have adopted. He sent the proposal to Demis Hassabis, CEO of GDM, who relayed it to senior policy staff. However, it went unanswered until Google signed an agreement.

Google’s leadership claimed it would not sign an agreement. The researcher disagreed, but his warnings were largely ignored. Although he managed to sow doubt within the Pentagon about the agreement, Google ultimately signed a contract handing over its AI with no restrictions on killer robots or mass AI surveillance. Google’s contractual restrictions were even weaker than OpenAI’s. At that point, he could no longer stay at Google in good conscience, so he decided to leave.

That is why Turner left Google DeepMind. It is also a broader story: how powerful individuals and institutions failed to honor their ethical promises under pressure.

Google and the Immigration Supply Chain

After the death of Alex Pretti, the researcher was determined to act effectively. To understand how to reduce the harm caused by the DHS, he examined the involvement of the big tech companies. Although Microsoft and Amazon have larger commitments, he was surprised to discover Google’s contracts with the DHS:

• The DHS 2025 AI use case inventory lists Google among the GenAI providers used to "improve operational efficiency" at the DHS.

• Google sells Cloud services to ICE through third parties such as ITC Federal.

• On October 3, 2025, Google removed applications that warned of ICE activity.

• Google voluntarily handed a protesting student’s account over to ICE without notice, breaking its terms-of-service promise to "notify the user by email before disclosing information [to the government]."

Faced with these findings, he wondered how to act effectively. The typical actions of activists, such as petitions, had already failed. Google had ignored a large petition on this issue. Moreover, Google’s leaders had probably hardened their company against traditional organizing tactics. Sit-ins, strikes, even a mass resignation of Google engineers: he considered all of them ineffective (if he could even carry them out).

On reflection, he concluded that Google would not care about 100 random research engineers resigning. In the AI industry, talent is concentrated and teams are run by a few stars who are hard to replace. He did not need to coordinate 100 engineers. Maybe he only needed to coordinate 10.

He had followed the news and guessed that Sundar Pichai (CEO of Google) was more of a businessman than someone who would say, "give me a big speech about ethics and I will change my mind." But if a few hard-to-replace people were ready to leave, it would matter to the company, so Sundar might listen.

That is when he remembered seeing Jeff Dean tweet about ICE’s bad reputation, retweeting quotes from Anne Frank. Maybe he did not even need 10 engineers, he just needed one.

Discussion with Jeff Dean

At first, he thought about who could put him in touch with Dean. But (and this is a good general lesson) if someone wants to talk to someone about something, they can always JUST ASK THEM!

He told Jeff that he respected him for speaking out, and that he wanted Google to withdraw from the DHS supply chain. He asked whether he shared these goals and, if so, how he could help.

Dean suggested it would be reasonable for the researcher to contact a few people. Their names: Sundar Pichai (CEO of Google), Demis Hassabis (CEO of Google DeepMind) and Thomas Kurian (CEO of Google Cloud). The researcher thought, "sure, Jeff. No problem. I’ll just tell them what I think."

The Email

The researcher writes as a concerned employee at GDM. He recently contacted Jeff Dean about his concerns. Dean suggested it would be reasonable to send an email to all three of them.

He has no problem with Google working with legal administrations of either American political party. His concern is not political, but rather the events facilitated by Google’s role in the DHS supply chain.

He believes ICE has far exceeded its legal mandate to remove illegal immigrants from the country in an orderly manner. According to observers, ICE operations frequently deprive targets of their right to a trial. These operations regularly detain citizens in facilities operating with little (or no) legal oversight. More than 1,000 people are reported missing at one such place. At other sites, the ACLU reports human rights abuses and a severe lack of safety for detainees.

These are not standard, legitimate enforcement activities. These operations are concerning from a human rights standpoint and also pose a reputational risk for any provider involved.

On January 28, 2026, the DHS published its 2025 AI use case inventory, which lists Google as one of several GenAI providers that "improve operational efficiency" at the DHS. The researcher urges Google to immediately stop working with ICE (and the DHS more broadly), including by supporting third-party integrators that facilitate these specific operations. Whether through a direct ("prime") contract or through third parties (such as ITC Federal), Cloud and Gemini must not fuel these operations.

History will judge the tech sector by its involvement in these events. The researcher loves working at Google, and he wants to make sure Google is on the right side of this story.

Alexander Matt Turner
Researcher, Google DeepMind

They never responded. The researcher went back to Jeff and asked him for lunch to discuss constructive opportunities for real change within Google. He told him: "anytime, anywhere. I’ll come to Mountain View to meet you."

At that point, he thought this was where "plan A" would fail. To his great surprise, Dean actually agreed, to a lunch a few weeks later.

A lot was going to happen by then.

The Pentagon Attempts to Intimidate Anthropic

On February 25, 2026, the Pentagon pressured the frontier AI lab Anthropic to remove the restrictions from its existing contract. These restrictions concerned lethal autonomous weapon systems and AI surveillance/profiling. The ultimatum was clear: "give us your product or we will designate you as a supply chain risk." The government wanted to use AI for "any lawful use."