Caitlin Kalinowski, who served as the head of hardware and robotics engineering at OpenAI starting in November 2024, has confirmed her resignation from the firm.
“I have stepped down from my position at OpenAI,” she stated. “I hold the Robotics team and our collective achievements in high regard, and this was a difficult decision. While AI plays a vital role in national security, the implementation of domestic surveillance lacking judicial review and the development of lethal autonomous systems without human oversight are boundaries that required more thorough consideration. My decision was rooted in principle. I maintain great respect for Sam and the rest of the team, and I am proud of the work we accomplished.”
Her exit occurs during a period of growing debate regarding the extent to which AI firms should assist the U.S. military. Recently, discussions between the Department of Defense and Anthropic ended without an agreement after Anthropic insisted on rigorous constraints regarding autonomous weaponry and domestic surveillance. Shortly thereafter, OpenAI secured a contract with the Pentagon to integrate its AI models into a secure government network.
This development faced backlash from various observers and staff members, who suggested that OpenAI moved to finalize the deal after Anthropic’s refusal. CEO Sam Altman subsequently admitted that the timing of the agreement appeared “opportunistic,” and the company has since worked to define clearer limitations on military applications of its technology.
An OpenAI representative verified her departure, stating: “We are confident that our partnership with the Pentagon establishes a responsible framework for national security applications of AI, while explicitly maintaining our red lines: a prohibition on autonomous weapons and domestic surveillance. We acknowledge the diverse perspectives on these matters and remain committed to ongoing dialogue with our employees, government officials, civil society, and the global community.”
Prior to joining OpenAI, Kalinowski spent nearly two and a half years as a hardware executive at Meta, where she spearheaded the development of Orion—formerly known as Project Nazare—which the company described as “the most advanced pair of AR glasses ever made.” Meta showcased the device in September.
Before her tenure on the Orion project, she spent over nine years developing virtual reality headsets at the Meta-owned company Oculus, following a nearly six-year stint at Apple, where she contributed to the design of MacBook Pro and Air models.