OpenAI reports that over 230 million people globally already ask healthcare-related questions on ChatGPT every week, making the launch of a dedicated health experience feel less like a surprise and more like inevitability. ChatGPT Health is positioned as a way for people to better understand elements of their care, including test results, appointments, diet and exercise, and medical insurance options. Additionally, in the United States, patients will be able to upload their private medical records and health app data to gain more insight into their health.

Crucially, OpenAI has emphasized that ChatGPT Health is not intended to replace medical care. Instead, it aims to support understanding, preparation, and engagement. Whether it fulfils that ambition or becomes a new source of friction will depend on how thoughtfully it is adopted.

From “Dr Google” to AI health partner

For patients, ChatGPT Health feels like the natural successor to “Dr Google.” Preparing for appointments, understanding results, and making sense of complex medical language have long been pain points – and they sit at the heart of ChatGPT Health’s promise.¹

AI has the potential to reduce barriers such as poor health literacy and fragmented access to health data, helping patients feel more informed and more in control of their care. However, we cannot ignore that AI has previously been implicated in misinformation in the healthcare space. ChatGPT has previously been found to provide unsafe advice for infections, and failed to ask clarifying questions. More recently, investigations into AI-generated health summaries by Google AI Overviews have revealed potentially dangerous inaccuracies.

ChatGPT Health’s core value, therefore, may not lie in giving answers, but in supporting better questions, better conversations, and better engagement with clinicians.

Healthcare’s new literacy challenge

National Academy of Medicine suggests that developing skills in AI health literacy (i.e. health literacy specific to AI-generated responses) will be paramount to meaningfully shift power toward patients by enabling them to explore treatment options, cross-check recommendations, and consider multiple perspectives before making decisions. Without AI health literacy skills, patients may misinterpret information, delay seeking care, or become anxious about benign findings.

There is an art to accurate prompting – without the correct context, AI could misdirect patients to incorrect guidelines, leading them down the wrong path when navigating their disease. This presents a new dynamic in healthcare. For years the healthcare sector has worked to improve patients’ health literacy, transforming complex science into simple, accessible language, running disease education campaigns, and providing innovative support programs. However, in this new era, prompt literacy is taking center stage. For biopharmaceutical companies, this creates an additional layer of support to consider.

A more informed patient or a more complex consultation?

For healthcare professionals, ChatGPT has the potential to be both an enabler and a source of new complexity. On the positive side, it helps patients arrive at appointments better prepared, which can support more focused consultations and shared decision-making.

On the downside, it may create extra work for time-strapped and overworked clinicians who must fact-check outputs, myth-bust misconceptions, and reassure anxious patients and caregivers. The confident tone of AI responses can also challenge professional authority when recommendations conflict with clinical judgement.

Only a day after OpenAI announced the rollout of ChatGPT Health, they also announced ChatGPT for Healthcare, designed to streamline workflows, summarize evidence, and support clinical reasoning within healthcare organisations.

The perceived benefit of ChatGPT for Healthcare is clear – it can reduce healthcare professionals’ time spent on documentation, diagnostics, retrieving treatment pathways, and administrative tasks, allowing clinicians to focus on direct care.

At the same time, such AI introduces governance, privacy, and bias risks, as both the patient and clinician-facing AI rely on sensitive health data. There is also the question of safety; although OpenAI promises ‘answers grounded in relevant medical sources,’ we must be wary of AI’s predisposition to ‘hallucinations’ and unsubstantiated content generation, hich could lead to misdiagnosis and misdirection to inappropriate medical advice. As a result, the use of AI by healthcare practitioners must always be underpinned with human experience and expertise.

Systemic efficiency gains and new risks

At a system level, ChatGPT presents significant opportunities. Enterprise AI tools such as ChatGPT for Healthcare can streamline operations, reduce administrative costs, and support population health initiatives.

Yet there are also potential risks. AI can increase health inequalities and contribute to overburdening healthcare systems – patients may demand further tests, second opinions or see a doctor for ultimately harmless ailments.

The challenge for healthcare systems will be to leverage AI’s efficiency gains while mitigating risks. Strong oversight, integration into established clinical workflows, and training for staff on the safe use of both patient-facing and enterprise AI will be essential to ensure ChatGPT enhances care rather than introducing new inefficiencies or disparities. 

Keeping people at its heart

AI will increasingly shape how healthcare is delivered, but people remain central to the system. The value of tools like ChatGPT Health and ChatGPT for Healthcare lies not in replacing human judgement, but in supporting patients, clinicians, and decision-makers to make better, more informed choices.

For patients, AI can improve understanding and engagement. For clinicians and health systems, it can enhance efficiency and insight. But AI has limitations. It is people – patients asking the right questions, clinicians applying experience and expertise, and organizations providing oversight – who ensure those tools are used safely and effectively.