Google Quietly Signs a Classified AI Deal with the Pentagon—Employees Are Furious

Google has reportedly signed a classified agreement with the US Department of Defense, giving the military access to its AI models for “any lawful government purpose.” The Information broke the story, and it landed less than 24 hours after a group of Google employees publicly demanded CEO Sundar Pichai block the Pentagon from using the company’s AI, citing fears it would be used in “inhumane or extremely harmful ways.”

If confirmed, this puts Google in the same boat as OpenAI and xAI, both of which have already inked classified AI deals with the US government. Anthropic was also on that list until it got blacklisted by the Pentagon for refusing to remove safety restrictions the DoD found inconvenient. So much for responsible AI.

What exactly does “any lawful government purpose” mean? That’s the part that should make everyone uneasy. It’s a broad, vague carve-out that could cover anything from logistics optimization to drone targeting. And since the deal is classified, we won’t get to see the fine print. The public gets a press release; the Pentagon gets the models.

Google’s own AI principles, published in 2018 after the infamous Project Maven backlash, explicitly ruled out building AI for weapons or surveillance that violates international norms. But principles are cheap when classified contracts come calling. The company has been quietly walking back those commitments for years, and this deal is the latest sign that the old guard has lost the internal debate.

What’s striking is the timing. Employees were literally in the middle of organizing against this exact scenario when the deal was reportedly signed. It suggests either a complete breakdown in internal communication or a deliberate decision to move forward regardless of employee concerns. Neither is a good look.

The Pentagon’s appetite for AI is voracious and bipartisan. Both OpenAI and xAI have already shown there’s no real reputational cost for these partnerships—at least not one that matters to the bottom line. Google’s calculus seems to be: if everyone else is doing it, why not us? It’s a cynical take, but not an inaccurate one.

Still, this feels different. Google has spent years cultivating an image of responsible AI development. It has ethics boards, public principles, and a CEO who talks about putting a “higher bar” on AI deployment. This deal makes all of that look like theater. When the classified contracts are signed and the employees are ignored, the principles don’t mean much.

I’m not naive—military contracts are lucrative and Google is a business. But the company could have drawn a line. It could have said no, like Anthropic tried to do. Instead, it chose to join the club. The question now is whether the employees who care about this will stay and fight or leave for companies that haven’t yet made their peace with the defense industry.

For now, the deal is done. The Pentagon gets its AI. Google gets its money. And the rest of us get to wonder what “any lawful government purpose” really means when it’s stamped classified.