OpenAI has unveiled a new ChatGPT feature in the United States that can review users’ medical records to deliver more tailored health-related responses, a move that has prompted privacy concerns from digital rights advocates.
The company is encouraging users to share medical records alongside data from fitness and wellbeing apps such as MyFitnessPal. This information is analysed to offer personalised insights and guidance.
OpenAI said conversations held within ChatGPT Health would be stored separately from other chats and would not be used to train its AI models. It also stressed that the feature is not intended for medical diagnosis or treatment.
Andrew Crawford from the US non-profit Center for Democracy and Technology said it was “crucial” that robust, “airtight” protections are in place when handling sensitive health information.
There is currently no confirmation on whether or when the feature will be rolled out in the UK.
“New AI health tools have the potential to empower patients and improve health outcomes, but health data is among the most sensitive information people can share and it must be safeguarded,” Crawford said. He added that AI companies are increasingly pushing personalisation to increase the value of their services.
“As OpenAI begins exploring advertising as part of its business model, it’s especially important that health data is kept strictly separate from other memories ChatGPT may retain from user conversations,” he said.
OpenAI says more than 230 million people worldwide ask its chatbot questions related to health and wellbeing each week. In a blog post, the company said ChatGPT Health includes “enhanced privacy protections” to safeguard sensitive data.
Users can choose to share information from platforms such as Apple Health, Peloton and MyFitnessPal, as well as upload medical records, allowing ChatGPT to provide more relevant answers to health-related questions. OpenAI emphasised that the feature is designed to support medical care, not replace healthcare professionals.
