OpenAI launched "Health in ChatGPT," a feature that integrates Apple Health, medical records, and wellness apps into the platform. The rollout targets U.S. users first.
The health feature creates a two-tier system. Paying subscribers access GPT-5.6 Sol, OpenAI's more advanced model. Free users get GPT-5.5 Instant, a weaker version. This distinction matters because health advice quality varies between models.
Over 300 million people ask ChatGPT health questions weekly. OpenAI recognizes that health guidance carries real stakes. Better models produce more nuanced, accurate medical information. Worse models risk harmful oversimplifications or missed context.
The tiered approach reflects OpenAI's business model. Premium subscriptions fund development. Free tiers drive adoption. But health advice sits in an awkward middle ground. Restricting quality health information to paying users creates equity problems. People with less money get worse health guidance exactly when they need it most.
The integration with Apple Health and medical records adds value for those who pay. ChatGPT gains context from users' actual health data, workout history, and medical documents. This personalization improves recommendations substantially. Free users lose this advantage entirely.
OpenAI doesn't explicitly market this as a medical tool. They position it as health information assistance, not diagnosis. But the practical impact matters. Free users receive genuinely inferior health information. The gap between GPT-5.6 Sol and GPT-5.5 Instant isn't marginal. One model handles complexity better than the other.
This pricing strategy raises regulatory questions. Medical advice quality impacts public health outcomes. Regulators may scrutinize whether paywalling health information violates consumer protection standards. OpenAI operates in a gray zone where their product influences health decisions without being classified as a medical device