OpenAI recently rolled out ChatGPT Health despite poor evidence of efficacy.

Disclaimers and caveats are likely to fall on deaf ears.

Industry leaders continue to call for greater restraint; but it may be too late for public health.

More than 300 million people consult ChatGPT for health-related questions every week. To accommodate the demand, OpenAI quietly rolled out ChatGPT Health last month to a select group of US adults; a soft opening of sorts. For the world’s first patient-focused AI, there was surprisingly little fanfare, and I can’t help but think there was a reason.

ChatGPT Health is based on the same underlying architecture as regular ChatGPT, but with a key difference: with permission, the AI can link to your medical records and download health metrics recorded by Apple wearables, such as the Apple Watch. Though initially a separate tab in ChatGPT’s interface, it’s now seamlessly integrated; an operational decision by OpenAI after seeing users posing health questions to the chatbot organically during daily conversations. An OpenAI press release said ChatGPT Health can “help explain a visit note, understand lab results, dig into doctor-patient discussions, and help prepare questions for follow-up appointments…” The platform can optimize the user’s experience with more “personalized conversations.”

But this initiative has tension at its heart. On the one hand, OpenAI qualifies the rollout by saying it’s not for diagnosis or treatment and is “not designed to replace the care and judgment of qualified medical professionals.” On the other hand, that’s exactly what the public will use it for, and OpenAI knows it. "The reality is, we're living in a time when not everyone has access to quality care or care in a timely manner,” said one executive to reporters last month. “The average doctor appointment in the United States is less than 15 minutes... You're often left on your own to sort through a lot of fragmented information.” For many, the black box that is ChatGPT........

© Psychology Today