© ROOT-NATION.com - Use of content is permitted with a backlink.
Google has signed a classified artificial intelligence agreement with the US Department of Defense that allows the Pentagon to use its AI models for “any legitimate government purpose.” This is evidenced by the sources of The Information.
This information comes just a few hours after an open letter to CEO Sundar Pichai was published calling for the abandonment of such classified military agreements in the AI field. It was signed by more than 560 Google employees. So far, the company has not confirmed or commented on the deal.
According to The Information, the agreement does not contain ethical restrictions similar to those that Anthropic included in its contract with the Pentagon. It was these restrictions that led to the company’s “characterization” as a national security risk in February 2026 and its blacklisting by the Donald Trump administration.
Read also: Trump vs. Claude AI: How the AI Debate Is Unfolding in the U.S.
Anthropic refused to lift the ban on mass domestic surveillance and fully autonomous weapons without human intervention. Instead, Google’s agreement reportedly allows the use of technology for “any legitimate government purpose” without such exceptions. This approach is in line with the no-restrictions model favored by the Trump administration, in contrast to the compromise agreed with OpenAI, which provides for certain “red lines” on domestic surveillance.
Currently, the Pentagon has signed classified AI agreements with four major US companies: OpenAI, xAI, Google, and, before being blacklisted, Anthropic. The latter was removed from the list of suppliers due to ethical restrictions. OpenAI revised the terms of the contract, keeping some of the restrictions. xAI signed an agreement with no apparent restrictions, and now Google has signed a contract with wording that probably gives the Pentagon the widest possible scope.
As a result, a closed pool of AI vendors has been formed, from which Anthropic is excluded, while the other three companies have different but significant levels of freedom to provide technology for military applications.
The timing of the employees’ letter is particularly telling. The very next day after it was published, the employees found themselves in a company that signed the very deal they were calling for the company to reject. This creates a direct and uncomfortable contrast that is likely to be discussed both in internal meetings and during press conferences.
Read also: How Steve Jobs Might Have Reacted to Gemini on macOS: Google “Reinvents” the Window
Google has never officially disclosed the detailed terms of its AI cooperation with the Pentagon, and the phrase “any legitimate government purpose” is based on information from one anonymous source cited by The Information.
The letter from the employees and the agreement with the Pentagon together outline the dilemma faced by all major AI companies. On the one hand, there is the U.S. government’s demand for unrestricted use of AI for military purposes. On the other hand, there are ethical principles that companies declared after the Project Maven scandal in 2018, including the avoidance of creating autonomous weapons without human control.
Anthropic chose to comply with the principles and was excluded from the contracts. OpenAI and Google seem to have chosen to participate in government contracts. Whether this decision will be temporary or final depends on the political situation and whether employees can influence the company’s position.
Read also:
- Google Translate turns 20 years old: Interesting facts, tips and new features
- Google Pixel overheats and melts: Users raise alarm over fresh update