OpenAI claims ChatGPT’s new default model hallucinates way less

OpenAI has released a new default model for ChatGPT, GPT-5.5 Instant, which significantly reduces hallucinations and inaccurate claims. This is a major improvement in factuality, especially on high-stakes topics like medicine, law, and finance. Users can expect fewer made-up responses from the model. This change matters as it increases the reliability of AI-generated information. No action is required from users, as the new model is the default.