A swift consumer backlash has rattled OpenAI’s flagship app, sending U.S. uninstalls of ChatGPT soaring 295% day-over-day on Saturday, February 28.
This came after news broke that the company had struck a deal with the U.S. Department of Defense, rebranded under President Donald Trump’s administration as the Department of War.
The spike, according to data from Sensor Tower, stands in stark contrast to ChatGPT’s average 9% day-over-day uninstall rate over the past 30 days. The surge signals not just momentary outrage, but a coordinated consumer reaction at scale — one that materially disrupted the app’s growth trajectory within 48 hours.
Register for Tekedia Mini-MBA edition 19 (Feb 9 – May 2, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab.
At the same time, rival AI firm Anthropic emerged as an unexpected beneficiary.
U.S. downloads of Anthropic’s Claude app rose 37% day-over-day on Friday, February 27, and climbed another 51% on Saturday, Sensor Tower reported.
Anthropic had publicly declined to pursue a partnership with the defense department, citing concerns that AI tools could be used for domestic surveillance or deployed in fully autonomous weapons systems that the technology is not yet equipped to handle safely.
The contrast between the two companies’ positions appears to have sharpened consumer choice. In the span of a weekend, what had largely been a competitive race over model performance and feature sets turned into a referendum on AI ethics and military alignment.
The data show a dramatic pivot in ChatGPT’s growth pattern. Before the deal became public, U.S. downloads had risen 14% day-over-day on Friday. By Saturday, downloads had fallen 13%, followed by a further 5% decline on Sunday.
Appfigures reported that on Saturday, Claude’s total daily U.S. downloads surpassed those of ChatGPT for the first time. Its estimates suggest Claude’s U.S. installs jumped 88% day-over-day. By the weekend, Claude had climbed to the No. 1 spot on the U.S. App Store, a gain of more than 20 positions compared with February 22.
Claude also became the No. 1 free iPhone app in six other countries — Belgium, Canada, Germany, Luxembourg, Norway, and Switzerland — underscoring that the surge was not confined to American users.
Similarweb added longer-term context, noting that Claude’s U.S. downloads over the past week were roughly 20 times January levels. The firm cautioned that not all of the growth can be directly attributed to political controversy, as broader product adoption trends may also be contributing.
Still, the timing is difficult to ignore.
Consumer response extended beyond downloads. Sensor Tower reported that one-star reviews for ChatGPT jumped 775% on Saturday and rose another 100% day-over-day on Sunday. Five-star reviews dropped by 50% over the same period. Such swings can have algorithmic consequences: app store rankings are influenced not only by installs but also by engagement velocity and review sentiment.
In effect, the backlash touched all major levers of app distribution — installs, uninstalls, ratings, and visibility.
The Economics of Defense Partnerships
The controversy exposes a structural dilemma facing AI developers. Training and deploying frontier AI models requires massive capital outlays for compute infrastructure, data center expansion, and specialized chips. Government contracts, particularly with defense agencies, offer stable, large-scale revenue streams that can underwrite these costs.
Yet consumer-facing AI products operate in a different trust economy. Millions of users interact daily with ChatGPT for writing, research, coding, and personal assistance. That user base includes educators, students, creatives, and professionals who may view defense collaboration through a different ethical lens.
While the tension between enterprise-scale funding and consumer trust is not unique to OpenAI, this episode has made it visible in real time.
OpenAI CEO Sam Altman acknowledged missteps in how the agreement was rolled out. In a post on Monday, he said the company “shouldn’t have rushed” the announcement and admitted, “I think it just looked opportunistic and sloppy.”
He wrote that the company had been “genuinely trying to de-escalate things and avoid a much worse outcome” and outlined plans to revise elements of the agreement.
His remarks suggest that, beyond the substance of the deal, the communication strategy played a central role in shaping public reaction.
Anthropic, which has emphasized safety and constraints around military applications, appears to have capitalized on the moment without directly attacking its rival. The company said it could not agree to terms that might allow its AI systems to be used for domestic surveillance or fully autonomous weapons.
The download surge gives Anthropic a rare consumer spotlight in a market where ChatGPT has long dominated mindshare. But the durability of that momentum will depend on retention, product performance, and continued alignment with user expectations.
Historically, app store spikes driven by controversy can fade as attention shifts. However, the magnitude of the weekend’s movement — particularly the 295% uninstall rate for ChatGPT — suggests this was more than routine churn.
A Politicized AI Marketplace
The events have exposed a broader shift: AI platforms are no longer evaluated solely on model accuracy, speed, or multimodal capability. Users are increasingly scrutinizing governance structures, political affiliations, and deployment contexts.
As generative AI becomes embedded in education systems, newsrooms, software development pipelines, and everyday communication, public expectations are rising. Partnerships with military or intelligence agencies, once confined to defense contractors, now involve consumer-facing tech brands with global audiences.
In a sector defined by rapid iteration and capital intensity, the weekend’s data show that public sentiment can move just as quickly — and at scale.