Google signed a classified AI deal with the US Department of Defense, allowing the Pentagon to use its AI models for “any lawful government purpose,” according to a report from The Information. This agreement excludes the ethical restrictions that contributed to the blacklisting of Anthropic in February 2026.
The deal, reported shortly after over 560 Google employees urged CEO Sundar Pichai to reject such military arrangements, highlights a conflict between employee concerns and government demands for unrestricted military AI capabilities. Google had not confirmed the specifics of the agreement at the time of the report.
Unlike Anthropic, which included prohibitions on mass domestic surveillance and autonomous weapons in its contract before being removed from the supplier pool, Google’s agreement does not impose such limitations. This aligns with the preferences of the Trump administration for broader use in military applications.
The Pentagon has now established classified AI contracts with four leading AI companies: OpenAI, xAI, Google, and Anthropic, which was recently blacklisted. Anthropic’s removal stemmed from its refusal to abandon ethical restrictions. In contrast, OpenAI renegotiated to retain some limits, while xAI signed an agreement without clear restrictions.
The juxtaposition of the employee letter and Google’s military deal poses significant implications for the company’s internal culture and its public stance on AI ethics. Pichai’s responses to employee concerns will be closely scrutinized in future engagements, particularly given the optic of their letter coinciding with the announcement of the classified deal.
The “any lawful government purpose” clause, cited from an anonymous source, frames a narrative that could create tension within the company. The ongoing situation underscores the broader challenge faced by AI firms in balancing operational needs with established ethical principles in the evolving political landscape. The choice made by Google and OpenAI appears to yield significant discretion for military applications, diverging from Anthropic’s ethical commitments.
Future dynamics will likely depend on political shifts and internal employee advocacy, especially among the signatories of the letter and their allies, who may influence internal decision-making regarding military AI engagements.
