OpenAI makes default ChatGPT more personal

OpenAI has updated ChatGPT's default model to GPT-5.5 Instant, making personalization a more visible part of the product.

In an official announcement published on May 5, OpenAI said the new default model is more accurate, more concise and better at using context from past chats, files and connected Gmail when helpful. The company said it is also designed to produce fewer hallucinated claims than the previous Instant model.

OpenAI introduced memory-source controls so users can see and manage the context ChatGPT uses for personalization. That gives users more visibility into how the system shapes responses.

TechCrunch, Axios, The Verge and 9to5Mac all reported the same rollout the same day, confirming that this is now a live product change rather than a preview.

The immediate question is how quickly the new behavior reaches all users and whether independent testing confirms OpenAI's accuracy claims. For now, the update signals a clear shift toward a ChatGPT experience that is more tailored to individual users while trying to stay tighter and more factual.