OpenAI bet that users would not care about defense contracts. The mass uninstalls proved otherwise. ChatGPT saw a 295% spike in uninstalls on February 28, 2026—the day after OpenAI announced its Pentagon deal—far exceeding its typical 9% day-over-day uninstall rate. One-star reviews jumped 775% while five-star reviews dropped 50%. The consumer and defense markets are fundamentally incompatible for a single AI brand.
OpenAI Bet That Users Would Not Care About Defense Contracts; the Mass Uninstalls Show That Consumer and Defense Markets Are Fundamentally Incompatible
TechCrunch reported that ChatGPT experienced a dramatic 295% spike in uninstalls on Saturday, February 28, 2026. U.S. downloads fell 13% that same day and continued declining 5% on Sunday, reversing the previous day’s 14% growth. The Indian Express noted that one tracking website claimed over 1.5 million users canceled their ChatGPT subscriptions in response to the deal. Forbes had reported the controversy triggered backlash among ChatGPT users who deleted the popular AI application in waves.
Anthropic’s Claude benefited directly from the backlash. As TechCrunch documented, Claude saw U.S. downloads surge 37% on February 27 and 51% on February 28 after Anthropic publicly refused to partner with the Department of Defense, citing concerns about AI being used for surveillance of Americans and fully autonomous weapons. Claude reached the number one position on the U.S. App Store on Saturday and became the top free iPhone app in six countries internationally. The CXO Digitalpulse analysis framed it as a clear consumer choice: users who objected to defense contracts voted with their feet.
OpenAI announced its Pentagon partnership shortly after Anthropic terminated similar negotiations. The contrast was stark: Anthropic refused terms that would allow domestic surveillance or fully autonomous weapons and was designated a supply-chain risk. OpenAI signed. The #QuitGPT movement emerged as users uninstalled ChatGPT and switched to Claude. Yahoo Finance reported the scale of the exodus—1.5 million users—underscoring that this was not a fringe reaction but a mainstream consumer response.
The lesson is structural. A brand that serves both consumers and the Pentagon faces an irreconcilable tension. Consumers who object to military AI will not stick around. Defense contractors and agencies will not accept the same restrictions that consumer users might prefer. OpenAI tried to have both. The uninstalls proved it cannot.
What This Actually Means
Consumer AI and defense AI cannot coexist in one brand. OpenAI learned that the hard way. The companies that thrive will either serve consumers with clear ethical boundaries or serve the Pentagon with “any lawful use”—but not both under the same name.