OpenAI has announced it will no longer remove older models without prior warning, following user backlash over the replacement of its 4o model with GPT-5. Nick Turley, OpenAI’s head of ChatGPT, acknowledged the oversight, stating, “In retrospect, not continuing to offer 4o, at least in the interim, was a miss.”
Turley expressed surprise at the “level of attachment” users had to the 4o model, noting that the strong feelings often related to the model’s “personality.” OpenAI is now actively working to integrate the “warmth” of 4o into GPT-5. CEO Sam Altman also commented on the update, hoping it would be “not as annoying (to most users) as GPT-4o.”
The initial decision to remove 4o was driven by a desire to simplify model choices for ChatGPT’s vast user base, which includes 700 million weekly users. Turley emphasized that the move was not financially motivated, stating, “It definitely wasn’t a cost thing.” Instead, the primary goal was to enhance simplicity, as the need for average users to select between different models was considered “cognitively overwhelming.” This policy change indicates OpenAI’s commitment to providing greater predictability for users regarding major model updates.
