Caitlin Kalinowski, who served as the head of robotics and consumer hardware at OpenAI, announced her resignation on Saturday, expressing concerns about the company’s recent agreement with the United States Department of Defense.
In a message shared on the social media platform X, Kalinowski said the company moved too quickly in approving the plan to deploy its artificial intelligence models on the Pentagon’s classified cloud infrastructure. According to her, the decision required deeper discussion and stronger safeguards before being finalized.
Kalinowski acknowledged that artificial intelligence can play a meaningful role in strengthening national security. However, she raised serious concerns about how the technology might be used. She warned that potential domestic surveillance of Americans without proper judicial oversight and the possibility of autonomous systems making lethal decisions without human approval are issues that require far greater scrutiny and debate.
Although she was not immediately available for direct comment, Kalinowski clarified in her post that she continues to hold strong respect for OpenAI’s leadership, including CEO Sam Altman, as well as the company’s broader team. Despite that respect, she argued that the announcement of the Pentagon partnership came before clear governance rules and ethical boundaries had been fully defined.
In a follow-up post, Kalinowski emphasized that her concerns were primarily related to governance and decision-making processes. She suggested that matters involving national security and advanced AI technology are too significant to be rushed through agreements or public announcements.
In response, OpenAI stated that the deal includes additional safeguards designed to ensure responsible use of its AI systems. The company reiterated that it maintains firm restrictions, including a ban on using its technology for domestic surveillance or fully autonomous weapons.
OpenAI also said it understands that opinions on AI and military collaboration can vary widely. The company pledged to continue engaging with employees, government institutions, civil society groups, and communities worldwide to address these concerns.
Kalinowski joined OpenAI in 2024 after previously leading augmented reality hardware development at Meta Platforms.
Also Read: OpenAI Questioned About its Safety Rules
