ChatGPT's Default Model Is Now "GPT-5.5 Instant" — Far Fewer Errors in Sensitive Areas

What happened: OpenAI updated ChatGPT's default model to "GPT-5.5 Instant." According to the announcement, mistaken assertions on topics requiring care—such as medicine, law, and money—have been roughly halved compared to before.

Why it matters: It's rolling out to free users in phases as well, switching to the new model automatically without any special action. It effectively raises the baseline quality of everyday use.

What it means: For fact-checking and research in Japanese, it's now easier to use with greater confidence than before. That said, for important information, keep verifying against official, primary sources.