The resignation of Caitlin Kalinowski from OpenAI, announced on March 7, 2026, highlights the deep tensions stirred by the contract with the Pentagon within the company and beyond.
A Leader Who Chooses Her Words Carefully
Caitlin Kalinowski, head of hardware and robotics at OpenAI, announced her departure in a post on X on March 7, 2026. Her profile places her among the most experienced technical executives in the industry: after nearly six years at Apple, where she contributed to the design of MacBooks, she spent over a decade at Meta, where she led Project Orion, an advanced augmented reality glasses initiative that has yet to be commercialized. She had joined OpenAI in 2024 to lead the hardware unit within the robotics division.
Her statement leaves no room for doubt. « I resigned from OpenAI. I have a profound commitment to the Robotics team and the collaborative efforts we undertook. This was not an easy decision, » she wrote in a post on X. She then clarified the reasons for her choice: « AI has an important role in national security. But surveillance of Americans without judicial oversight and lethal autonomy without human authorization are lines that deserved more deliberation than they got. »
An OpenAI spokesperson confirmed the resignation to TechCrunch and defended the contract: « We believe our agreement with the Pentagon creates a workable path for responsible national security uses of AI while making clear our red lines: no domestic surveillance and no autonomous weapons. »
An Agreement Signed in a Tense Political Climate
The contract between OpenAI and the US Department of Defense was announced on February 27, 2026. « Tonight, we reached an agreement with the Department of War to deploy our models in their classified network, » Sam Altman, CEO of OpenAI, wrote in a post on X. According to Altman, the agreement includes « prohibitions on mass domestic surveillance and human accountability in the use of force, including for autonomous weapons systems, » and the Pentagon « agrees to these principles, reflects them in its law and policy, and we have inscribed them in our agreement. »
The timing of the announcement is particularly telling. The day before, the US government had declared Anthropic, OpenAI’s main rival, a risk to the national supply chain. This classification followed the refusal of Dario Amodei, CEO of Anthropic, to allow its AIs to facilitate mass surveillance or integrate into autonomous weapons systems. OpenAI thus seized the commercial and political opportunity created by its competitor’s exclusion.
But the speed with which the agreement was finalized is precisely what Kalinowski and other employees criticize. According to information reported by Business Insider, internal discussions on the ethical implications of certain military uses were reportedly not concluded before the deal was publicly announced.
Employees in Rebellion, Users Fleeing
Kalinowski is not the only one to have expressed disagreement or left the company. Aidan McLaughlin, a researcher at OpenAI, publicly regretted the agreement, while another employee confided to Business Insider that his colleagues admired Anthropic for resisting the Pentagon’s demands. Clive Chan, a technical team member, wrote on X that he was advocating for more transparency on the contract’s terms, specifying: « If we later learn this is not the case, I will advocate internally to terminate the contract. »
The reaction of ChatGPT users was equally decisive. On February 28, 2026, the day after the announcement, the daily uninstall rate of the application surged by 295% compared to its usual average, according to Sensor Tower data reported by Business Insider. Reddit posts massively called on users to « cancel ChatGPT. » At the same time, Claude, Anthropic’s chatbot, climbed to the top of the most downloaded free apps on the US App Store, a position it held several days after the announcement.
Faced with the controversy, Sam Altman announced that he had amended the contract’s text to better prevent the most sensitive uses. The BBC reported that OpenAI now describes its agreement with the military as an « agreement with more guardrails than any prior agreement for classified networks. » For many observers, however, this modification comes after the fact and appears more like a response to public pressure than a genuine fundamental revision.
OpenAI Caught Between Commercial Ambition and Ethical Credibility
The agreement with the Pentagon represents a major commercial victory for OpenAI, securing direct access to the US defense markets in a context where its direct competitor has been excluded. But it illustrates a fundamental tension in the company’s positioning: the search for lucrative government contracts versus the safety principles that have long constituted OpenAI’s public trust argument.
Caitlin Kalinowski’s departure provides a concrete measure of this unease. A technical executive with a profile spanning Apple, Meta, and OpenAI chose to leave rather than endorse an agreement she deemed insufficiently considered. For OpenAI, the issue is no longer just about defending its positions with the Pentagon: it is now about regaining the trust of those who run its labs, and of the millions of users who are beginning to look at the competition with renewed interest.