OpenAI has announced ChatGPT Health, a new service that promises to consolidate users’ medical records and health and fitness apps, such as Apple Fitness and MyFitnessPal, in one place, alongside their conversations with the AI assistant on health-related topics.
The company emphasizes that ChatGPT Health is not intended for diagnosis or treatment, but this is unlikely to alleviate concerns regarding the growing use of artificial intelligence by the general public, the trend toward lighter regulation in the US, and documented cases of poor-quality advice leading to patient harm.
According to OpenAI, more than 230 million people worldwide ask health and wellness-related questions in ChatGPT every week. The company also published a report earlier this week stating that a quarter of Americans use ChatGPT for health inquiries at a time when public opinion of the US healthcare system is at its lowest level in 24 years.
For now, the service is being presented as a preview, with OpenAI opening a waitlist for people who wish to try it before its official release to all users on web and iOS versions in the coming weeks. However, this list is not available to people in Europe, likely reflecting the difficulties associated with complying with the region’s stricter data protection requirements.
According to OpenAI, ChatGPT Health can be used for tasks such as interpreting medical test results, preparing for doctor appointments, receiving nutrition and exercise advice, and comparing different health insurance options based on individual health status.
The company also states that health information and inquiries will be separated from other conversations with the AI assistant and will be protected by additional multi-layered security measures to maintain confidentiality. During implementation, OpenAI worked with the digital company b.well Connected Health, including on the integration with electronic health records.
In response to data security concerns, OpenAI has committed to not using users’ health information to train its language models.
Nevertheless, experts advise individuals to carefully consider whether to share medical and health information with a chatbot. Numerous uncertainties remain, including where and for how long the data will be stored, whether it will be reviewed by human operators, and whether data control mechanisms are sufficiently reliable and transparent to prevent misuse or hacking attacks.
Ashley Alexander, Head of Health Products at OpenAI, stated that people are already using ChatGPT for health questions to better navigate situations where the healthcare system does not fully meet their needs. According to her, health information is fragmented across various systems, apps, and trackers, and the average doctor’s visit lasts less than 15 minutes, leaving long periods between appointments during which people want to take a more active role in understanding and managing their own health.
OpenAI adds that the feature will have core protection mechanisms to ensure users have real control over their data. This includes temporary chats, the ability to delete conversations from OpenAI’s systems within 30 days, and training models not to retain personal information from user conversations. The company emphasizes that the goal is to provide reliable support that assists, rather than replaces, healthcare professionals.
READ
MORE
