What Happened
Caitlin Kalinowski, OpenAI's head of robotics and consumer hardware, resigned on March 7 over the company's agreement with the Department of Defense. She had led hardware and robotics engineering at OpenAI since November 2024, after stints at Meta building the Orion AR glasses and nearly a decade at Oculus.
In a post on X, Kalinowski wrote that "surveillance of Americans without judicial oversight and lethal autonomy without human authorization are lines that deserved more deliberation than they got." Her core objection was not necessarily the deal itself, but how fast OpenAI moved to close it.
The timeline matters here. On February 27, Anthropic's negotiations with the Pentagon collapsed after the company refused to drop a clause preventing its models from being used for bulk data analysis on Americans. Hours later, the Trump administration blacklisted Anthropic, and OpenAI announced its own deal to deploy models on Pentagon classified cloud networks. Sam Altman himself admitted the deal was "definitely rushed" and that "the optics don't look good."
OpenAI says the contract includes red lines: no domestic mass surveillance, no autonomous weapons without human control, and no high-stakes AI decisions without human approval. But Kalinowski clearly felt those assurances were insufficient, or at least insufficiently deliberated.
Why It Matters
This is the highest-profile departure from OpenAI directly tied to the Pentagon deal, and it hits a sensitive area. Kalinowski was not some random engineer. She was running the robotics division, the part of the company most directly connected to physical-world AI applications. When your robotics lead quits over concerns about autonomous weapons, that is a signal worth paying attention to.
For anyone choosing between AI platforms, the Pentagon situation has created a real philosophical split. Anthropic pushed for strict guardrails and got blacklisted. OpenAI moved fast and closed the deal. CNN reported internal staff were "fuming" about the arrangement. Meanwhile, Anthropic's Claude jumped past ChatGPT in the Apple App Store the day after the news broke. Users are voting with their downloads.
If you use ChatGPT or Claude for work, this does not change your day-to-day workflows. But it tells you something about the companies behind the tools you depend on and what they will prioritize when principle and opportunity collide.
Our Take
The Pentagon was always going to get AI. That was never in question. The question was whether AI companies would negotiate hard on the terms or race to sign first. OpenAI chose speed. Anthropic chose conditions. And the market is reacting.
Kalinowski's resignation is not an isolated event. It is part of a pattern at OpenAI where senior people leave when the company's direction shifts in ways they did not sign up for. The list of departed safety and ethics-focused staff keeps growing.
For tool users, here is the practical takeaway: if data governance and ethical guardrails matter to your organization, you should be tracking these deals closely. The company that builds your AI stack is making decisions about military deployment, surveillance boundaries, and autonomous systems. Those decisions reflect the same judgment they apply to your data and your use cases.
We would not tell anyone to switch tools over this alone. ChatGPT is still a strong product. But when your robotics chief resigns saying the company did not think hard enough before signing a military deal, that is worth factoring into your long-term platform decisions.