Despite the active discontent of hundreds of its employees, Google has entered into an agreement with the Pentagon for access to the AI tools of Gemini. This deal has sparked intense debates, especially against the backdrop of similar contracts with OpenAI and Anthropic's refusal of such cooperation.
According to the obtained data, this agreement imposes on Google the obligation to adjust the settings and security filters of its AI models at the government's request. Notably, the contract clearly states: "The AI system shall not be used for mass surveillance within the country or as autonomous weaponry without proper human oversight."
However, Google has no right to control or block the government's decisions regarding the use of this AI assistant. Pentagon representatives, for their part, preferred to refrain from commenting on this information.
It is worth noting that OpenAI already entered into a similar deal on February 27, 2026, despite opposition from 96 employees who signed a corresponding letter. This demonstrates a general trend in the collaboration of tech giants with the defense department.
Unlike others, Anthropic firmly rejected such an agreement, insisting on solid guarantees that their AI model Claude would not be used for surveillance. It is important to emphasize that at that time, a clause similar to the current one in the deal with Google was absent from the proposed contract.
Leave a comment