OpenAI Expands ChatGPT Health to All US Users
· news
OpenAI Makes ChatGPT Health Available to All US Users
OpenAI’s decision to expand its ChatGPT Health feature to all US users over 18 has sparked concerns about liability for medical decisions. The company is walking a tightrope between providing valuable information and avoiding legal repercussions.
The scale of user engagement with health-related queries is significant, with 300 million interactions per week since the feature’s introduction in January. This surge in usage raises questions about the accuracy and safety of AI-powered health advice. While OpenAI’s models have improved in answering health-related questions, as demonstrated by GPT 5.6-Luna outperforming its predecessor on the HealthBench evaluation, this progress does not necessarily translate to safer medical guidance.
OpenAI maintains that its services are not intended for diagnosing or treating medical conditions and is working to make health-related answers more reliable. However, several studies have highlighted the limitations of AI bots in providing accurate medical advice, emphasizing the need for users to verify information and consult professionals.
The rollout of ChatGPT Health marks a critical juncture for AI in healthcare. As companies like Anthropic and Google launch their own health-related features, OpenAI’s approach highlights the tension between accessibility and accountability. The company’s decision to make ChatGPT Health available to all US users underscores the need for transparency and accountability while providing accessible information.
OpenAI emphasizes verifying information and taking professional advice, a prudent step forward in mitigating risks associated with AI-powered health advice. However, it remains to be seen whether this approach will be sufficient to ensure safety and accuracy. The stakes are high: get it right, and AI could revolutionize healthcare; get it wrong, and the consequences could be severe.
The future of AI-powered health advice hangs precariously in the balance as OpenAI’s approach sets a new standard for companies to follow. As this space unfolds, one thing is clear: the need for transparency, accountability, and accurate information will continue to drive innovation and reform in healthcare.
Reader Views
- CMColumnist M. Reid · opinion columnist
While OpenAI's expansion of ChatGPT Health to all US users is a step forward in accessible healthcare information, we mustn't lose sight of the elephant in the room: data security. With 300 million interactions per week, the potential for sensitive medical information to be compromised or misused is alarmingly high. OpenAI needs to prioritize robust data encryption and secure user authentication to prevent unauthorized access. This is not a trivial concern – as AI-powered health advice becomes increasingly prevalent, the risks of data breaches and identity theft are only set to escalate.
- RJReporter J. Avery · staff reporter
As OpenAI expands ChatGPT Health to all US users, we're witnessing a classic case of technology outpacing regulatory frameworks. While the company's emphasis on verification and professional advice is welcome, it glosses over the fact that AI models can only be as good as their training data – and health data is notoriously messy. The real concern isn't just accuracy, but also who bears liability when users act on faulty or incomplete information. OpenAI needs to engage in a more nuanced conversation about accountability, rather than simply shifting responsibility back onto consumers.
- CSCorrespondent S. Tan · field correspondent
"The real challenge for OpenAI and other AI healthcare providers lies in defining the boundaries of their liability. As ChatGPT Health expands its reach, users will increasingly rely on AI-generated advice, blurring the lines between consultation and treatment. What's missing from this discussion is a clear framework for holding these companies accountable when their guidance leads to adverse outcomes. Without stricter regulations, we risk creating a healthcare landscape where AI-driven advice supplants human expertise."