I lived through Google’s AI-military crisis. Here’s why engagement still matters

Years before Anthropic’s recent public clash with the U.S. military, I found myself at the center of a similar crisis. At Google, an AI contract with the Pentagon as part of its Project Maven program spiraled into a public scandal about tech collaboration with the military. Google ended up canceling the contract — the result of a controversy that was shaped as much by fear of AI and misperception as by the facts. Nearly 10 years later, the questions and dynamics are resurfacing — though this time with much higher stakes.

AI has advanced since the Pentagon initiated the 2018 Maven project. The systems being debated today don’t just identify objects in drone footage — they reason, plan, and act. The central question then and now is the same: Should AI companies engage with the military, and on whose terms? The answer matters more than ever — because the companies most worried about how this technology could be misused may be the ones the military most needs at the table.

I…