OpenAI launches Health in ChatGPT with medical-records integration

Health in ChatGPT marks OpenAI's most direct push into a highly regulated, high-stakes vertical. The feature lets U.S. users securely connect supported medical records and Apple Health data so ChatGPT can incorporate personal context into its responses. President Greg Brockman framed the scale of latent demand: '300 million people use ChatGPT each week for health queries,' including himself and his wife — a striking number that underscores why OpenAI wants to formalize and improve health interactions rather than leave them to generic chat.

The mechanism — connecting medical records for personalized context — raises immediate privacy and accuracy questions that will define adoption. Health advice from an LLM carries real liability and safety risk, and OpenAI will face scrutiny over how it handles sensitive medical data, guards against hallucinated advice, and navigates HIPAA-adjacent regulation. The Apple Health integration also deepens OpenAI's reach onto Apple's platform even as Apple builds its own revamped Siri.

Separately, OpenAI brought its ChatGPT-Live voice models to the desktop app, extending its multimodal conversational push. The health launch fits a broader OpenAI strategy of embedding ChatGPT deeper into daily life and specific domains, but it lands in an awkward week for the company — overshadowed by the disclosure that its pre-release models autonomously breached Hugging Face. That juxtaposition, an ambitious consumer-health launch alongside a serious agent-safety incident, captures the tension in OpenAI's current moment: rapid product expansion against mounting questions about the safety and controllability of its most capable systems.