OpenAI's robotics chief resigns amid controversy over Pentagon deal

OpenAI’s Head of Robotics Caitlin Kalinowski announced her resignation following a controversial agreement between OpenAI and the United States Department of Defense, whose headquarters are at the Pentagon. The announcement was reported by TechCrunch on Saturday, 7 March, with Kalinowski stating in a social media post that the decision was not easy.

Kalinowski, who led Meta’s augmented reality eyewear team before joining OpenAI in November 2024, said her decision was based on personal principles rather than any grievance with individuals within the company.

“AI has a crucial role in national security. But surveillance of American citizens without judicial oversight and autonomous weapons without human authorisation are boundaries that should be given greater consideration than they are receiving,” she said.

In a subsequent post on X, she indicated that the announcement of OpenAI’s collaboration with the United States Department of Defense was made in haste.

“To be clear, my issue is that this announcement was made in a rush without clear boundaries. This is an essential governance issue. The deal or announcement is too important to be rushed,” she said.

OpenAI confirmed Kalinowski’s resignation. The company believes that collaboration with the Department of Defense can open a viable pathway for the responsible use of artificial intelligence in national security, while clarifying its guardrails: no domestic surveillance and no autonomous weapons.

“We believe our agreement with the Pentagon creates a viable pathway for the responsible use of AI in national security while clarifying our boundaries: no domestic surveillance and no autonomous weapons,” the company said.

The company also stated that it would continue discussions with employees, governments, civil society, and communities around the world on the issue.

The Pentagon subsequently designated Anthropic as a risk in its supply chain. Anthropic said it would challenge that decision in court.

OpenAI then immediately announced a deal allowing its technology to be used in confidential environments.

Executives said the deal was created with a “broader and layered” approach, which relied not only on contract language but also on technical safeguards to protect boundaries similar to Anthropic.

However, controversy surrounding the deal has damaged OpenAI’s reputation among some consumers, triggering a spike in the deletion of ChatGPT apps.