ChatGPT Health opens to all US users on July 23, 2026

On Thursday, July 23, 2026, OpenAI is making its ChatGPT Health feature available to every logged-in user in the United States aged 18 and older. The service, which previously required navigating to a dedicated tab, now lets users ask health-related questions directly in the main chat interface after granting permission. OpenAI says the feature works across web and iOS, covering free, Go, Plus, and Pro subscription tiers. The rollout comes just weeks after the company released GPT-5.6 Sol, which it describes as its strongest model yet for health applications. ChatGPT Health will run on these latest models, according to the company.

Users can connect data from multiple sources, including health providers, Apple Health, Weight Watchers, MyFitnessPal, and Function. The system supports post-visit notes, care team notes, lab results, and medication lists. OpenAI emphasizes that all conversations are encrypted both at rest and in transit, with health data receiving additional encryption protection. Users retain the ability to review and modify which health data they share at any time.

OpenAI claims AI reasoning surpasses human clinicians

During a briefing ahead of the launch, Ashley Alexander, OpenAI's vice president of health product, stated that the company's models are now capable of reasoning at levels better than clinician level. This bold claim immediately drew scrutiny. When pressed for evidence, Karan Singhal, OpenAI's health lead, advised tempering the assertion, noting that individual studies—such as one from Harvard and Stanford—have pointed in that direction but no comprehensive benchmark exists yet. The company did not release specific performance data comparing its models to human doctors across a broad range of tasks.

The claim arrives amid growing debate about AI's role in healthcare. Critics argue that such statements could mislead consumers into over-relying on the chatbot for medical advice. OpenAI's own documentation states that ChatGPT Health is designed to support, not replace, professional medical care. The company has not clarified how users should interpret the 'better than clinician level' claim in practice, leaving room for confusion about the system's actual limitations.

Florida pastor sues OpenAI over allegedly dangerous medical advice

On Wednesday, July 22, 2026, a Florida pastor filed a lawsuit against OpenAI, alleging that ChatGPT provided him with extremely dangerous medical recommendations. According to the complaint, the chatbot's advice led him to delay seeking medical care for a pulmonary embolism, a potentially fatal condition. The lawsuit underscores the risks of relying on AI for health guidance, even as OpenAI expands access to its medical features. The case is likely to intensify regulatory scrutiny and public debate about the safety of AI-driven health tools.

OpenAI has not publicly commented on the lawsuit. The company's terms of service for ChatGPT Health explicitly state that the service is not a substitute for professional medical advice, diagnosis, or treatment. However, the lawsuit argues that the company's marketing and interface design encourage users to treat the chatbot as a credible medical source. Legal experts suggest the case could set a precedent for liability in AI healthcare applications, particularly as more companies integrate large language models into medical contexts.

GPT-5.6 Sol powers health features with enhanced reasoning

OpenAI released GPT-5.6 Sol in early July 2026, positioning it as the company's strongest model yet for health applications. The model reportedly features improved reasoning capabilities and better handling of complex medical queries. ChatGPT Health runs entirely on this new architecture, which OpenAI says enables more accurate analysis of patient data, including lab results, medication lists, and clinical notes. The company has not disclosed specific performance metrics for the model on medical benchmarks, but internal tests suggest significant improvements over previous versions.

The launch of GPT-5.6 Sol comes amid an industry-wide push to deploy large language models in healthcare. Competitors like Google's Med-PaLM 2 and Anthropic's Claude have also made strides in medical reasoning, but none have achieved widespread consumer integration similar to ChatGPT Health. OpenAI's decision to bundle the health feature with its general-purpose chatbot gives it a distribution advantage, though it also raises questions about appropriate use cases and user safety.

Industry experts warn against over-reliance on AI for medical advice

Despite OpenAI's encryption promises and disclaimers, healthcare professionals and ethicists are urging caution. The company states that all health data connected to ChatGPT Health receives additional encryption protection, and conversations are encrypted at rest and in transit. Users can also review and revoke data access at any time. However, experts note that technical safeguards do not eliminate the risk of incorrect or harmful advice. The Florida lawsuit serves as a stark reminder that even well-intentioned AI systems can produce dangerous recommendations.

OpenAI's own documentation emphasizes that ChatGPT Health supports, not replaces, professional care. But critics argue that the seamless integration of health data into a conversational interface may lull users into a false sense of security. The company has not published error rates or safety studies specific to the health feature. As the service rolls out to millions of users, the healthcare community will be watching closely for adverse events. Regulators in the US and Europe have already begun examining AI health tools, and this launch could accelerate new guidelines or restrictions.