Google has signed a classified agreement with the U.S. Department of Defense allowing the Pentagon to deploy its Gemini AI models for “any lawful government purpose,” according to The Information, which first reported the deal Tuesday citing a person familiar with the matter.
The agreement places Google alongside OpenAI and Elon Musk’s xAI, both of which have their own classified AI deals with the U.S. military. Classified networks handle a range of sensitive defense operations, including mission planning and weapons targeting.
The deal expands an earlier arrangement. Google had signed an initial agreement in November 2025 limiting Gemini to unclassified government work. The new amendment opens access to classified systems, a significant escalation in the company’s defense footprint.
What the Contract Says — And What It Doesn’t
According to The Information’s reporting, the contract includes language stating that Google’s AI “is not intended for, and should not be used for, domestic mass surveillance or autonomous weapons (including target selection) without appropriate human oversight and control.”
However, a separate clause in the same agreement states that it “does not confer any right to control or veto lawful Government operational decision-making” — meaning Google retains no practical authority to block how the Pentagon actually deploys its models.
The contract also requires Google to help adjust its AI safety settings and content filters at the government’s request, a provision critics say effectively gives the military discretion to weaken the company’s own guardrails.
600 Employees Signed a Letter. Pichai Signed the Deal.
The timing of the disclosure was striking. Just one day before the deal was reported, more than 560 Google employees — many from the company’s AI-focused DeepMind division — sent an open letter to CEO Sundar Pichai urging him to block Pentagon access to Google’s AI, warning the technology could be used in “inhumane or extremely harmful ways.”
It is not the first time Google employees have pushed back on military AI work. In 2018, thousands of workers successfully pressured the company to let its Project Maven contract — which used AI to analyze drone footage over Iraq and Syria — lapse without renewal.
This time, the outcome was different.
The Anthropic Contrast
Google’s decision stands in sharp contrast to the path taken by Anthropic, the maker of the Claude AI model. Anthropic had signed a $200 million Pentagon contract in July 2025, becoming the first AI lab to integrate frontier models into classified defense workflows.
Renegotiations collapsed in February 2026 after Anthropic CEO Dario Amodei refused to accept the same “any lawful use” language now embedded in Google’s agreement. Amodei insisted on contractual prohibitions against using Claude for domestic mass surveillance or fully autonomous weapons systems without human oversight.
The Pentagon responded by designating Anthropic a supply chain risk — a label typically reserved for foreign adversary contractors — effectively barring Pentagon suppliers from using Claude in their work. President Trump followed with a directive ordering all federal agencies to immediately cease use of Anthropic’s technology.
Anthropic sued the Trump administration in March, and a federal district judge in San Francisco temporarily blocked enforcement of the blacklisting, calling the government’s actions “Orwellian.” A federal appeals court later declined to issue a stay, leaving Anthropic excluded from DoD contracts while litigation continues.
A Race to the Pentagon
OpenAI moved quickly after Anthropic’s blacklisting. OpenAI struck its own Pentagon deal within hours of Anthropic being designated a supply chain risk in February. xAI followed. Now Google has joined them.
The Pentagon signed agreements worth up to $200 million each with major AI labs in 2025, according to Reuters. For Google, the stakes extend beyond a single contract. The company’s Public Sector division has set an internal target of adding approximately $6 billion in contract value between 2025 and 2027, as it looks to close the gap with rivals Amazon Web Services and Microsoft Azure in the government cloud market.
Google and Alphabet Did Not Comment
Alphabet and the Department of Defense — officially renamed the Department of War by President Trump — did not immediately respond to requests for comment at the time of publication. A spokesperson for Google Public Sector confirmed to The Information that the new agreement is an amendment to an existing contract, without elaborating on its terms.
The classified nature of the deal means the full scope of Gemini’s intended military applications will not be made public.