Sam Altman Signs Pentagon Deal as QuitGPT Boycott Hits 4 Million Participants

Sam Altman signed an agreement with the US Department of Defense on the very day his main competitor publicly refused to do the same, and the move triggered one of the most high-profile reputational crises in tech industry history. In February 2026, Anthropic's CEO said he could not in good conscience give the US military unrestricted access to the company's AI systems. The position was clear and stated publicly: AI in the hands of the military without ethical constraints is too great a risk, and no commercial interest outweighs responsibility to society.

It was a rare case of a major tech company saying "no" to a government customer. Washington responded quickly. The Trump administration labeled Anthropic a "supply chain risk", a designation usually reserved for unreliable foreign contractors.

The political pressure was as heavy as it gets: the company was effectively told that its refusal would have consequences. That was the moment Altman signed the contract. He accepted what his competitor had turned down.

The OpenAI chief later admitted the deal "looked opportunistic and sloppy", but the admission came too late and did nothing to change how it was already perceived. The internet reacted fast. ChatGPT uninstalls rose 295% that same day.

QuitGPT, a movement urging users to drop OpenAI products, gathered 2.5 million participants in its first week, and by the time of writing it had reached 4 million. People explain their choice simply: they don't want their data and money going to a company that handed its systems' capabilities to the military without any public discussion of ethical frameworks.

Against this backdrop, Anthropic's Claude took the No. 1 spot among free apps in the US App Store. Users looking for an alternative found one quickly, and they chose the company whose leader had publicly put principles ahead of commercial gain.

The crisis spread beyond the user base. OpenAI's head of robotics publicly announced his departure, citing the Pentagon contract directly. Hundreds of employees signed an open letter supporting Anthropic's position.

One of the world's most influential tech companies saw an open split between its management and part of its team. Graffiti reading "you suck" appeared on the sidewalk outside OpenAI's San Francisco headquarters. It's an unusual turn for a company that until recently pitched itself as building AI for the benefit of all humanity.

This is a rare case of one company's ethical refusal instantly becoming another company's competitive advantage. Anthropic ran no ad campaigns. Its CEO simply said "no", and the market responded.

QuitGPT is not a fringe movement of a few hundred activists. Four million people deliberately giving up a free, convenient product on ethical grounds sends a signal: part of the audience is willing to accept inconvenience for the sense of having made the right choice. The movement will most likely not show up on OpenAI's revenue charts in the short term.

ChatGPT remains the dominant product, and enterprise customers make decisions based on functionality and price, not sidewalk graffiti. But reputational damage works differently from financial damage. OpenAI spent years building a narrative around responsible AI, a commitment to safety, and a mission to benefit humanity, and all of it collapsed within a few hours of a single deal.

Competitors didn't have to do anything: all it took was saying "no" at the right moment.