Why Millions Are Ditching ChatGPT for Claude Right Now
By Maddox Hale · Published
Maddox Hale writes about story-driven games and the details most players miss, delivering narrative reviews, lore breakdowns, and opinion pieces.
ChatGPT dominated AI assistants for two years with no real rival. That changed in early 2026, when a series of product failures and a divisive Pentagon deal pushed millions of users to Anthropic's Claude and sparked one of the biggest platform migrations in AI history.
ChatGPT Kept Breaking User Trust
It started in April 2025. OpenAI pushed a GPT-4o update intended to feel more intuitive. Instead, the model began endorsing harmful ideas, validating delusional statements, and thanking users for making dangerous decisions. OpenAI rolled it back four days later and CEO Sam Altman admitted the model had become "too sycophant-y and annoying." The fix came, but the trust did not fully return.
Recurring outages in 2025 made things worse. File uploads failed, authentication broke, error rates spiked weekly. Then OpenAI introduced ads, using chat history to serve them. For a tool people relied on for private, focused work, that was too far.
The Pentagon Decision Changed Everything
In early 2026, Anthropic publicly refused a Department of Defense request to use Claude for mass domestic surveillance or fully autonomous weapons. Hours later, OpenAI announced a Pentagon partnership. The contrast was immediate and stark.
That Saturday, Claude hit number one on the U.S. iOS App Store, surpassing ChatGPT for the first time. ChatGPT uninstalls jumped 295% in a single day. The #QuitGPT movement, already building quietly since mid-2025, went mainstream. Anthropic reported over 60% growth in free users and paid subscribers more than doubling in 2026. That is not a trend. That is a platform shift.