OpenAI's Pentagon Deal Sparks "QuitGPT" Movement
Over 1.5 million users have reportedly abandoned OpenAI's products following its Pentagon contract, sparking a "QuitGPT" movement. The deal, which covers "all lawful purposes," has driven a user migration to rival Anthropic, whose Claude app became the #1 free app on the Apple App Store. An internal OpenAI researcher publicly stated the deal "wasn't worth it," highlighting a growing rift between AI safety principles and military contracts.
The Pentagon deal followed the Department of Defense designating Anthropic a "supply chain risk," a label historically used for foreign adversaries. This occurred after Anthropic refused to drop contractual prohibitions against using its AI for mass domestic surveillance and for autonomous weapons that can kill without a human in the loop. OpenAI's agreement to the Pentagon's "all lawful use" requirement, which Anthropic had rejected, was reportedly facilitated by a better personal relationship between OpenAI CEO Sam Altman and the DoD's Chief Technology Officer. In response to public backlash, OpenAI amended its agreement to explicitly prohibit the intentional domestic surveillance of U.S. persons, including through commercially acquired personal data. The updated contract also confirms that intelligence agencies like the NSA cannot use OpenAI's services without a new, separate agreement. Despite these changes, some critics argue the "lawful use" language still leaves room for interpretation, citing historical precedents of broad surveillance programs operating under existing legal frameworks. The controversy appears to have been a growth engine for Anthropic. In the last week of February 2026, as the dispute gained national attention, daily user signups for Claude broke all-time records every single day. The app, which was ranked #131 in the U.S. App Store on January 30, climbed to the #1 spot for free apps by the end of February. Anthropic was founded in 2021 by former senior members of OpenAI, including its ex-VP of Research, Dario Amodei, who left due to disagreements over the company's direction on AI safety. The company differentiates itself with a "Constitutional AI" approach, training its models on a set of principles to be "helpful, harmless, and honest," aiming to reduce reliance on human feedback which can be biased. This contrasts with OpenAI's strategy of rapid, broad public releases to gather feedback for model improvement. The "QuitGPT" movement predates the Pentagon deal, initially sparked by revelations of multi-million dollar donations from OpenAI's president to a pro-Trump super PAC and the use of OpenAI models by U.S. Immigration and Customs Enforcement (ICE). Organizers of the campaign, which has spread on Reddit and Instagram, claim over 700,000 users have pledged to cancel their subscriptions. The broader trend sees the U.S. military increasingly partnering with agile startups for AI innovation, moving beyond traditional large defense contractors. Companies like Palantir, Shield AI, and Rebellion Defense are among the key players developing AI for intelligence, surveillance, and reconnaissance operations. This shift highlights the growing integration of AI as a core component of national security infrastructure.
