OpenAI's Healthcare Push: Karan Singhal Leads AI Health Research

OpenAI is significantly expanding its focus on healthcare, with Karan Singhal leading these health research initiatives. More than 230 million people use ChatGPT weekly for health and wellness advice. Singhal's goal is to make the chatbot more accurate, useful, and trusted by both patients and healthcare professionals. He previously helped develop medical AI models at Google. The GPT-5 family of models is the first set at OpenAI trained with a stronger emphasis on health-related performance. To improve ChatGPT's healthcare capabilities, Singhal built a team of health researchers and partnered with over 200 physicians. This collaboration led to HealthBench, a framework for evaluating AI health performance. OpenAI's latest model, GPT-5.5 Instant, performed better than physician responses and GPT-4o in company evaluations. They also saw a 71% reduction in flagged inaccurate health responses. OpenAI introduced a feature allowing users to connect health apps and upload medical records, which helps the AI understand personal medical history. This matters because it shows how AI is evolving to provide more personalized and reliable health information.