A 24-hour doctor may be available for you right in your smartphone with Artificial Intelligence (AI) fully entrenched in our everyday lives. But, the question of whether it can ever replace human doctors is still a big debate.
OpenAI launched ChatGPT Health, a new health and wellness experience within the ChatGPT platform made to help users better understand and manage their health. This new upgrade aims to bring personalised, data-informed insights directly to users while promising strong privacy protections and a supportive role alongside when it comes to getting medical advice.
In this guide, we look at what ChatGPT Health actually is, how it works, and check whether AI can realistically replace doctors at any level now or in the future. Before beginning, it is important to point out that ChatGPT Health has just launched, so research on it is very limited.
What is ChatGPT Health? Full Guide
ChatGPT Health is basically a new feature from OpenAI that creates a different health-focused space within the ChatGPT interface that is normally used, allowing users to ask medical and wellness questions in a specific environment built for that.
As per reports, it is designed to let people securely connect their medical records, wearable and wellness apps such as Apple Health, Function and MyFitnessPal, and use that information to get personalised answers and insights about any medical issues they have. Unlike standard ChatGPT chats, the health space reportedly uses special protections to secure sensitive data and keeps health related conversations separate from other chats, it seems. These conversations and any connected data are not used to train OpenAI's foundation models, which is something OpenAI has clearly said.
Moreover, the launch is a response to the huge volume of health related queries already sent to ChatGPT by everyday users. According to OpenAI, over 230 million people worldwide ask questions about health and wellness on the platform every week. ChatGPT Health plans to make these interactions more relevant and contextual by using a user's own health data where authorised, helping them interpret lab results, prepare for doctor visits, understand diet and exercise factors, and even compare health insurance options.
Furthermore, the product was developed with input from more than 260 practising physicians across 60 countries, with feedback on model outputs more than 600,000 times to fine tune responses and ensure clarity and safety. OpenAI has also created a bespoke evaluation framework called HealthBench, which uses physician written standards to assess answers for safety, appropriate escalation of care, and clinical relevance. Importantly, OpenAI says that ChatGPT Health is not designed to diagnose or treat medical conditions, but to support users in navigating everyday health questions and preparing for professional consultation.
Also, privacy and security are the main parts of the design. ChatGPT Health operates in a separate space from regular chats, with purpose built encryption and clear user controls over what is shared. Users can connect or disconnect apps and medical records at any time, and all data stays compartmentalised to protect sensitive information. Although some features like medical record integration are presently limited to certain regions such as the United States, full roll-out to web and iOS platforms is planned soon. You can join the waitlist here.
Read More: 'Not Kidding' : Elon Musk Warns Twitter Users on Misusing AI
Read More: 'This is Not Therapy': Microsoft AI CEO Reacts to People Using AI for Breakups, Family Problems
Can AI Realistically Replace Doctors at Any Level
Now, despite the tech jumps represented by ChatGPT Health, the idea that AI will soon replace doctors entirely is not supported by current evidence or expert opinion. AI tools, including ChatGPT, are undeniably powerful for processing and organising information, offering general guidance, and summarising complex medical data. Research shows that many healthcare professionals already use AI tools to assist with administrative tasks, summarise consultations, or support decision making. For example, a study in the UK found that nearly 30% of general practitioners are using AI tools during patient consultations for tasks including diagnostics support and administrative work, although worries exist about safety, privacy and regulation. There is some precedent here, as just a short time ago, Elon Musk's xAI Grok was credited with saving a man's life when doctors did not originally see a disease in him. So people have indeed benefited from AI health advice.
However, limitations obviously exist as well. A systematic review and meta analysis of ChatGPT's medical responses found an overall accuracy of around 56% across diverse query types, highlighting variability and significant concerns about reliability and consistency. AI systems often lack the comprehensive clinical context and nuanced judgement that experienced clinicians bring to diagnoses and treatment plans. These tools can also fail in providing emotional support and empathy, which are essential aspects of patient care that contribute to satisfaction and adherence to medical advice, so hallucinations can still exist, and the human element is still missing. But these studies were before the launch of ChatGPT Health, so how that will change these studies is yet to be seen.
