Improving health intelligence in ChatGPT

OpenAI has released GPT-5.5 Instant, a model update designed to improve the accuracy and reasoning capabilities of ChatGPT’s health-related responses. The company reports a 71 percent reduction in error rates and claims the model outperformed physician-written answers in internal testing for clarity and completeness. By utilizing physician-informed evaluations to refine its output, this update signals a shift toward deploying large language models for more reliable medical information. Users may now receive more technically precise guidance, though the technology remains a supplement rather than a substitute for professional care.