The Pentagon controversy has claimed its first high-profile departure at OpenAI: Caitlin Kalinowski, head of OpenAI’s hardware and robotics division, left the company on Saturday. Not for a better offer — on principle.
What Happened
Kalinowski joined OpenAI in November 2024 and was building out the robotics and hardware division. In a public statement, she wrote that she stepped down “on principle.” Her specific complaint: the policy guardrails around certain AI applications were not sufficiently defined before OpenAI announced the Pentagon deal.
Her words: “Surveillance of Americans without judicial oversight and lethal autonomy without human authorization are lines that deserved more deliberation than they got.”
The Context
OpenAI’s Pentagon contract came together after negotiations between the Defense Department and Anthropic fell apart. Anthropic had insisted on ethical guardrails — no mass surveillance, no autonomous weapons. The Pentagon apparently accepted these conditions from OpenAI without issue — or the question was never raised.
Kalinowski isn’t the first voice at OpenAI to raise concerns. But she’s the most senior person to leave over them.
Why This Matters
High-profile departures over ethical concerns send a signal — both internally and externally. When a leader responsible for one of the company’s most strategically important areas quits on principle, it raises questions.
The timing is awkward for OpenAI. The company is trying to position itself as a responsible Pentagon partner. A resignation over exactly those issues doesn’t make that easier.
My Take
What impresses me: Kalinowski isn’t leaving quietly. She’s naming the problems publicly and making clear where her line is. In an industry where most departures come with diplomatic platitudes, that’s refreshing — and brave.
Sources: