Remember Project Maven in 2018? Thousands of Google employees protested an AI partnership with the US military — and Google pulled out. In 2026, things are different.
What happened
Google signed an agreement with the Pentagon allowing Gemini to be deployed in the Department of Defense’s classified networks — ‘for any lawful purpose.’ Analysis, logistics, large-scale data processing. Google is one of eight tech giants that have signed such deals. OpenAI, Microsoft, Amazon, Oracle, Nvidia, and SpaceX are also on board.
Notably absent: Anthropic. They remain on the Pentagon’s blacklist.
The protest
Close to 600 Google employees signed an open letter criticizing the decision. But unlike 2018, there are no mass resignations, no public crisis. The power dynamics have shifted: in a market where AI engineers are still in demand but Meta and Microsoft just cut 20,000 jobs, protesting comes with more personal risk.
The bigger question
The Pentagon has brought eight companies on board — everyone except Anthropic. This makes the question ‘Should AI be used in the military?’ increasingly academic. The real question becomes: Which AI will be used there? And who sets the rules?
My take
Google once experienced Project Maven as a PR disaster. Now they’re signing the deal while everyone else does it too — there’s safety in numbers. I respect the 600 employees who signed the letter. But realistically: when OpenAI, Microsoft, and even SpaceX are at the table, a single protest isn’t going to change the trajectory.
Sources: