OpenAI Pentagon Deal Triggers Resignation, Boycotts, and Congressional Pushback

Summary

OpenAI finalized a deal with the Pentagon to deploy its AI systems across US national security operations. The contract triggered an immediate wave of consequences: the resignation of Caitlin Kalinowski, who had led the company's robotics division; a record spike in ChatGPT uninstallations; in-person protests outside OpenAI's San Francisco headquarters; and a legislative proposal in the US Congress. CEO Sam Altman publicly acknowledged that the announcement process had been rushed.

The episode deepens the divide within the AI industry over military and surveillance use cases, and throws into sharp relief the contrasting stance taken by Anthropic, which had previously refused a similar deal with the Pentagon — and suffered political consequences for it.

In practice

Caitlin Kalinowski, who led OpenAI's robotics division, announced her resignation on social media on March 7. In her statement, she acknowledged that AI has a legitimate role in national security but drew a line at two specific points: surveillance of Americans without judicial oversight and lethal autonomy without human authorization. She described the problem as "a governance concern first and foremost" and criticized the lack of sufficient deliberation before the agreement was struck and announced.

OpenAI confirmed Kalinowski's departure and defended the contract, stating it contains safeguards limiting how its technology may be used. The company said its models cannot be used for mass domestic surveillance, autonomous weapons systems, or high-stakes automated decision-making without human oversight. Deployment is cloud-only, which OpenAI says prevents its models from being used directly in physical autonomous weapons. Engineers with security clearances will remain involved in the deployment process.

On the user side, data from Sensor Tower showed ChatGPT uninstallations rising more than 295 percent on February 28, the day after the deal was made public. At the same time, Anthropic's Claude climbed to the top spot among free apps on Apple's US App Store, overtaking ChatGPT, and also became the most downloaded productivity app on the platform. Activists launched a protest campaign called "QuitGPT" with demonstrations outside OpenAI's headquarters.

In Congress, California Democratic Representative Sam Liccardo introduced an amendment to the Defense Production Act aimed at preventing the Pentagon from retaliating against companies that impose safety restrictions on high-risk technologies. Liccardo argued that companies building powerful AI should have a say in how their technology is deployed. The amendment failed, however, on a 16-25 vote.

Facing mounting criticism, Sam Altman acknowledged publicly on X that "the process was definitely rushed, and the optics don't look good." In an internal memo made public, Altman also said the contract language had been revised to include clearer prohibitions against domestic surveillance using "commercially acquired" personal data, and reiterated that OpenAI's models cannot be directed at autonomous weapons systems.

Who benefits / who loses

The deal places Anthropic and OpenAI in visibly opposing camps. Before the OpenAI-Pentagon deal, it had been Anthropic negotiating a similar partnership — but CEO Dario Amodei refused to allow the company's technology to be used for domestic mass surveillance or military attacks without human control. The standoff resulted in President Donald Trump ordering federal agencies to stop using Anthropic technology, and the Defense Department labeling Anthropic a "supply chain risk" — a designation typically applied to entities tied to foreign adversaries.

OpenAI moved into the space vacated by Anthropic and signed its own deal with the Pentagon. The controversy intensified further when the US and Israel launched military strikes against Iran shortly after the agreement was finalized.

In the short term, Anthropic appears to be the commercial beneficiary of its public positioning: Claude's jump to the top of app store rankings suggests a user migration driven by rejection of OpenAI's stance.

Why it matters

• The case shows that AI governance decisions in a defense context now carry immediate and measurable reputational and commercial consequences for companies.
• The resignation of a senior leader on principle signals growing internal tensions at frontier AI companies over the acceptable limits of their technology's use.
• The legislative response — even though it failed — indicates that the US Congress is beginning to engage in the debate over whether AI companies should have the right to define the terms of use of their systems in military contexts.
• The Anthropic vs. OpenAI dynamic around the Pentagon contract sets a precedent for what happens when an AI company says no to a government — and what happens when another says yes.