A leading executive at OpenAI has stepped down in response to the company’s recent partnership with the US Department of Defense, expressing concerns that the agreement lacked appropriate safeguards when it was unveiled.
Caitlin Kalinowski, who oversaw OpenAI’s robotics and consumer hardware division, announced her resignation via social media posts over the weekend. She had joined the organization in 2024, having previously managed augmented reality hardware projects at Meta.
Kalinowski stated on X that while AI plays a crucial role in national security, she believed that aspects such as surveillance of US citizens without judicial oversight and the deployment of lethal autonomous systems without human approval required more thorough consideration.
She clarified that her decision was based on principles rather than personal relationships, expressing admiration for CEO Sam Altman and the OpenAI team. In a subsequent post, she highlighted that her concern revolved around the lack of defined safeguards in the announcement process, stressing the importance of robust governance protocols.
OpenAI acknowledged Kalinowski’s resignation when contacted by TechCrunch. The company released a statement affirming that their collaboration with the Pentagon aimed to establish responsible applications of AI in national security, with clear boundaries prohibiting domestic surveillance and autonomous weaponry. OpenAI recognized the diverse opinions surrounding these issues and committed to ongoing dialogues with employees, governmental bodies, and civil organizations.
The partnership between OpenAI and the Pentagon was disclosed slightly more than a week ago, following failed negotiations between the Pentagon and rival AI company Anthropic. Anthropic had pushed for safeguards against widespread domestic surveillance and fully autonomous weapons, leading the Pentagon to label them as a supply-chain risk.
