OpenAI this week unveiled ChatGPT Health, a new feature within its artificial intelligence platform that aims to offer more informed and personalised answers on health and wellbeing topics. The service, however, is not a substitute for medical diagnosis or treatment, as the company itself stresses.
ChatGPT Health appears as a standalone tab, with its own history and memory, designed to ensure a safer and more private environment. OpenAI invites users to connect the chatbot to personal health data, namely medical records and apps such as Apple Health, Peloton, MyFitnessPal, Weight Watchers and Function, with the aim of generating answers with greater context.
Integration with healthcare providers
To make this integration possible, the American company has partnered with b.well, a platform that ensures technical compatibility with around 2.2 million healthcare providers. Through this link, ChatGPT will be able to analyse laboratory tests, consultation summaries and medical history, as well as wellbeing data relating to diet, physical activity, sleep and other vital indicators.
Beta version and limited access
The launch is taking place in experimental (beta) mode and only via a waiting list, and is still restricted to a small group of users. OpenAI guarantees, however, that access will be progressively extended to all users of the platform, regardless of subscription type.
Despite the promises of innovation, the company acknowledges the risks inherent in applying AI to sensitive subjects, warning of the need for responsible use. “ChatGPT Health is not a replacement for healthcare professionals”, OpenAI reiterates in an official statement, echoing earlier criticism aimed at the platform itself and competing products over inappropriate clinical advice.
A focus on reliability and safety
According to OpenAI, more than 230 million people use ChatGPT every week for health and wellbeing topics, many of them outside the conventional hours of medical services. Over the past two years, the company is said to have collaborated with more than 260 doctors from 60 countries, gathering feedback on more than 600,000 interactions with a view to improving the system.
Mental health was approached with particular care at this launch stage. Although the system is capable of handling delicate subjects, OpenAI says it is focused on directing users to professional or family support or suitable resources whenever it detects higher-risk situations.
Data protected, but with legal exceptions
On privacy, the company assures users that ChatGPT Health operates as a separate, encrypted space whose data is not used by default to train OpenAI's models. Even so, the company admits it may be legally required to provide information in cases of a court order or emergency.
As for American legislation, OpenAI clarifies that the HIPAA law – which protects health data in the US – does not apply to consumer products, such as ChatGPT Health, but only to formal medical settings.
The launch of this new feature represents a strategic step by OpenAI to strengthen the presence of artificial intelligence in users' daily lives, especially in an area where usefulness, trust and responsibility are constantly put to the test.
