Inside OpenAI’s Decision to Kill the AI Model That People Loved Too Much
February 12, 2026

(WSJ) – ChatGPT’s 4o model was beloved by many users, but controversial for its sycophancy and real-world harms linked to some conversations
Estrella is part of a vocal community of loyal 4o users who are in shock after OpenAI’s announcement in late January that it will retire the 4o model permanently on Feb. 13, saying its traffic had dwindled. The change means that paying ChatGPT users, who can pick which model they talk to, will have to select from other models that 4o fans say feel more distant.
The announcement signaled the end of the road for an AI model that proved sticky for users, helping drive OpenAI’s fast consumer growth and attracting a set of fans for whom it felt like a friend and confidant. But it has also been criticized for being overly sycophantic toward users, and doctors have linked it with cases of chatbot users developing psychotic delusions. (Read More)