An increasing number of patients are seeking health advice from AI chatbots, with nearly half of Americans turning to these digital tools for everything from lifestyle changes to cancer treatment opinions. However, experts warn that while these AI systems are designed with certain safeguards, they may still be influenced by external forces that shape the medical information you receive.
Dr. Isaac Kohane, the founding chair of the Department of Biomedical Informatics at Harvard Medical School and a coauthor of the book “The AI Revolution in Medicine: GPT-4 and Beyond,” emphasizes the potential risks associated with relying on AI for health decisions. He presents a scenario where a patient diagnosed with a brain tumor faces a recommendation for surgery based on standardized protocols, despite a more effective radiation treatment available at a specialized cancer center in the Midwest.
As healthcare systems increasingly adopt AI technologies, there is a growing concern that these systems may enforce a uniform standard of care that benefits financial interests rather than patients. With the U.S. healthcare industry valued at $5 trillion, the pressure to utilize AI for clinical decision-making could lead to an escalation in both unnecessary procedures and missed preventive measures.
Dr. Kohane notes that while standards of care are crucial, they can be interpreted differently by various clinicians. The shift towards AI-driven recommendations raises the question of whether patients and healthcare professionals will have the ability to challenge these recommendations. As AI becomes more embedded in healthcare practices, its influence could lead to a “monolithic” standard that limits options for patients and clinicians alike.
To navigate this evolving landscape, patients are encouraged to become informed consumers of AI-generated medical advice. Dr. Kohane suggests leveraging AI’s strengths, such as its capacity for providing multiple perspectives. Patients can ask the same health question from different angles or seek opinions from various chatbots, including Claude, ChatGPT, and Gemini. Research shows that these chatbots frequently present differing views on the same medical issues, highlighting the importance of obtaining diverse insights.
Moreover, patients should take ownership of their health data. The 21st Century Cures Act allows individuals access to digital versions of their health records, which can be stored in private folders. This data can then be utilized by AI systems to provide more personalized recommendations. However, caution is advised, as many chatbot companies do not guarantee that they will not retain or learn from users’ data.
In terms of policy, Dr. Kohane advocates for careful legislative measures to regulate the AI healthcare industry. He warns against hasty laws that could inadvertently favor established companies at the expense of innovative, patient-centric solutions. Effective regulation should focus on transparency rather than prescribing specific medical practices.
Patients should demand clarity on the origins and influences behind AI systems. Key questions include: What data was used to train the chatbot? Who influenced its clinical reasoning? How is patient data managed? Enhancing transparency could allow patients to make more informed decisions based on a clearer understanding of the AI’s recommendations and underlying principles.
As AI technologies continue to transform healthcare, the balance between serving patient needs and addressing profit motives will be crucial. The current moment presents an opportunity for patients to actively engage with their health data and question AI-generated guidance. By treating their health decisions with the same scrutiny as journalists do with sources, patients can advocate for their best interests and ensure that the evolution of healthcare technology aligns with their needs rather than those of a $5 trillion industry.
