Pew Research: US Public Demands AI Healthcare Patient Disclosure

• Over 70% of Americans believe doctors must disclose AI use in their healthcare, yet only 22% confidently understand it.
• Nearly a third of Americans have used AI chatbots for medical advice, with most finding them helpful, particularly younger adults.
• AI in healthcare spans from high-risk diagnostic tools requiring stringent FDA oversight to lower-risk administrative applications.
• Open-source AI models are not covered by HIPAA, posing significant data privacy risks for medical information.
• Experts emphasize that patient trust and broad AI adoption depend on clear disclosure, education, and explicit consent.

Over 70% of Americans consider it extremely or very important for their doctor to inform them if AI is being utilized in their treatment.

A recent survey by the Pew Research Center reveals a significant public expectation for explicit disclosure regarding the use of artificial intelligence in medical care. The findings indicate that over 70% of Americans consider it extremely or very important for their doctor to inform them if AI is being utilized in their treatment. This strong demand for transparency contrasts sharply with a general lack of understanding, as only 22% of respondents expressed confidence in their comprehension of how physicians employ AI, and nearly half (50%) were unsure if AI had even been a part of their healthcare experience.

Furthermore, the Pew AI healthcare patient disclosure report highlights a desire for greater patient agency, with more than half of those surveyed feeling they currently lack a say in their doctors' use of AI and wishing for more control over such decisions. This sentiment extends specifically to diagnostic applications, where over 80% of individuals want to be informed if AI is analyzing scans or assisting in diagnoses. Despite these concerns about AI integration in professional medical settings, the survey also points to a growing trend of individuals independently engaging with AI.

Nearly a third of Americans have turned to AI chatbots for medical advice or diagnosis, with the vast majority finding these tools helpful. This usage is particularly prevalent among younger adults aged 18-29, who are the most likely demographic to consult AI chatbots for health-related information, despite generally expressing more distrust in AI overall. Many of these users cited the free and affordable nature of chatbots as a primary motivator, with a smaller segment using them to better understand existing diagnoses or lab results, potentially reflecting the impact of rising medical costs and lower insurance rates among younger populations.

The term 'AI' in healthcare encompasses a much broader range of technologies than the generative AI models that have recently captured public attention, such as ChatGPT. As public health lecturer Tim Curry notes, machine learning has been researched and applied in medical practices for decades, with examples like RCG machines using basic AI for pattern interpretation. This distinction is crucial for understanding the varied implications of AI integration.

Experts differentiate between high-risk applications, such as AI embedded in diagnostic and treatment algorithms, and lower-risk administrative uses like scheduling and billing. While administrative AI can enhance operational efficiency, diagnostic AI carries significant risks. A critical concern highlighted by the report is that open-source AI models are not subject to HIPAA, potentially exposing users to substantial HIPAA AI data privacy risks when medical information is disclosed. This underscores the need for careful consideration of data handling and patient consent across all AI applications.

It is important to recognize that the data fueling these AI models often originates from conventional medical interactions, including appointments and testing. Therefore, the integration of AI is not a separate process but rather an extension of existing healthcare practices, necessitating a comprehensive approach to patient education and consent.

The findings from the Pew Research AI healthcare survey underscore a critical need for robust regulatory frameworks and clear ethical guidelines to foster trust in AI within the medical field. Jyoti Pathak, dean of Arizona State University’s School of Technology for Public Health, emphasizes that patients possess a fundamental right to be aware when AI influences their care, asserting that widespread adoption of healthcare AI hinges on trust, which in turn relies on transparent disclosure.

For high-risk AI applications, particularly those involved in diagnostics and treatment, stringent FDA oversight AI medical devices is deemed essential. This oversight must be complemented by rigorous clinical validation, explicit transparency in how AI operates, and thorough review by clinicians. Both Pathak and Curry advocate for comprehensive education and US AI healthcare patient consent mechanisms for all types of AI use, recognizing these as inseparable components of responsible AI integration.

Ultimately, the ethical integration of AI in healthcare demands a proactive approach to patient disclosure and consent. As AI becomes more commonplace, ensuring that patients are fully informed and have a voice in its application will be paramount for building and maintaining public trust, addressing AI medical ethics US concerns, and navigating the evolving landscape of digital health.

This Pew report highlights a critical public expectation for explicit disclosure and consent regarding AI use in healthcare. Lawyers and compliance officers should proactively review patient consent forms, data privacy policies (especially concerning non-HIPAA-covered AI tools), and monitor evolving regulatory guidance to mitigate legal and reputational risks associated with AI integration.

Source: Original reporting via Courthouse News