Caitlin Kalinowski, head of hardware and robotics at OpenAI (Unlisted:OPAI), has resigned following the company’s agreement to deploy its artificial intelligence models on US Department of Defense cloud networks.
Kalinowski, who joined OpenAI in 2024 after leading augmented-reality hardware projects at Meta Platforms, announced her resignation on March 7, citing concerns about governance and oversight.
In posts on X and LinkedIn, she said AI can play a role in national security but expressed unease over potential applications she described as “surveillance of Americans without judicial oversight” and “lethal autonomy without human authorization.”
She criticized the timing of the Pentagon deal, noting that it was announced before sufficient safety and oversight measures were established. “This was about principle, not people,” Kalinowski wrote.
I resigned from OpenAI. I care deeply about the Robotics team and the work we built together. This wasn’t an easy call. AI has an important role in national security. But surveillance of Americans without judicial oversight and lethal autonomy without human authorization are…
— Caitlin Kalinowski (@kalinowski007) March 7, 2026
OpenAI said the agreement provides a “workable path” for responsible military use of AI and maintains strict boundaries, including prohibitions on domestic surveillance and autonomous weapons systems.
The company highlighted that it will continue discussions with employees, government partners, and civil society on the responsible deployment of AI.
The Pentagon deal allows OpenAI’s models to be used on classified networks for military and domestic security applications. The announcement came shortly after competitor Anthropic reportedly declined Pentagon requests for unconstrained AI use, including fully autonomous weapons, and was labeled a potential “supply chain risk” by the US Department of War.