ChatGPT’s mobile app faced an unusually sharp consumer backlash after OpenAI announced a partnership with the U.S. Department of Defense, with U.S. uninstalls spiking 295% day over day, according to Sensor Tower. The surge far exceeds the app’s typical daily uninstall fluctuation of roughly 9%, signaling a notable shift in sentiment tied directly to the defense deal.
Data Shows Uninstalls Surged as Downloads Declined Sharply
Sensor Tower’s data points to a swift, measurable response: uninstalls spiked following the deal’s announcement, while U.S. downloads for ChatGPT dipped 13% the next day and slid another 5% the day after. Prior to the news, downloads had been trending up, including a 14% daily gain just before the announcement, highlighting how quickly consumer momentum reversed.
Ratings activity mirrored the downturn. Sensor Tower reports that 1-star reviews for ChatGPT jumped 775% on the day after the announcement and doubled again the following day, while 5-star reviews fell by 50%. The pattern suggests that a portion of the user base is expressing disapproval through direct app-store feedback rather than quiet churn alone.
Competitive Ripple Effects as Anthropic’s Claude Surges in Downloads
Rival Anthropic benefited from the moment. After publicly stating it would not partner with the defense department due to concerns about potential surveillance and autonomous weapon use, Anthropic’s Claude app saw downloads rise 37% one day and 51% the next, Sensor Tower estimates. Appfigures pegs the Saturday-over-Saturday jump even higher at 88%, and says Claude’s daily U.S. downloads surpassed ChatGPT’s for the first time.
Claude also climbed to the No. 1 spot on the U.S. App Store and topped free charts in several other markets, including Canada, Germany, Norway, Switzerland, Luxembourg, and Belgium, per Appfigures. Similarweb adds that Claude’s recent U.S. downloads were roughly 20 times higher than in January, though it cautions that broader factors, such as product updates and word of mouth, may also be at play.
Why the Backlash Accelerated After OpenAI’s Defense Deal
Defense work has long been a pressure point for major tech platforms. Past controversies around government AI projects, including Google’s Project Maven and debates over the Pentagon’s cloud initiatives, show how quickly public opinion can swing when consumer-facing brands venture into national security. The optics are particularly sensitive for AI assistants that live in people’s pockets, where trust and perceived neutrality are core to adoption.
Three issues appear to be driving the reaction. First, users worry about data governance: who can access prompts and outputs when military contracts are involved. Second, there’s fear of misuse in surveillance and autonomous systems, even if current agreements focus on nonlethal or back-office use cases. Third, consumers increasingly “vote” in app stores, where uninstalling or leaving a low rating is fast, visible, and cathartic—turning policy decisions into instant market signals.
Signals From the App Stores Show Shifting Trust and Churn
App store charts now double as real-time barometers of tech trust. The 295% surge in ChatGPT uninstalls alongside the spike in 1-star reviews is a stark indicator of reputational risk translating into measurable distribution headwinds. Meanwhile, Claude’s ascent to No. 1 underscores how quickly switching costs can erode in a market where top chatbots are broadly substitutable for casual use.
For OpenAI, the risk is not just one weekend’s numbers. App discovery and retention are compounding games: a few days of negative velocity can ripple through rankings, organic visibility, and word of mouth. For Anthropic, the test will be converting a values-driven bump into durable engagement and paid adoption, especially among enterprise users who care about uptime, privacy controls, and compliance as much as ethics statements.
What to Watch Next as Vendors Clarify Public-Sector Plans
Clarity on the scope of the OpenAI–DoD work will matter. Transparent limits on data access, deployment guardrails, human-in-the-loop requirements, and independent audits could calm consumer concerns. In parallel, watch whether web usage backfills any mobile softness; power users often pivot platforms rather than abandon tools entirely.
Finally, expect more explicit positioning from AI vendors on public-sector partnerships. As the market matures, ethics policies are becoming go-to-market levers. The latest download charts show that, at least for now, millions of consumers are paying attention—and are willing to act.
