Pew Survey: US Public Divided on AI Healthcare Decisions

• Americans are split on whether AI will improve patient health, with significant wariness about its impact on the doctor-patient relationship.
• The public is open to AI handling administrative tasks like scheduling and translation, especially when human oversight is involved.
• Support for AI collapses for high-stakes decisions such as cancer diagnoses or insurance claims, with most preferring no AI involvement or requiring human review.
• A majority of 60% would feel uncomfortable if their doctor relied on AI for diagnoses and treatments.
• Confidence in institutions like insurers and government to ensure AI is used in patients' best interest is low, though doctors fare better.

Public enthusiasm for artificial intelligence in healthcare diminishes sharply when the technology is proposed for high-stakes decisions, such as determining a cancer diagnosis or adjudicating insurance claims.

A recent Pew Research Center AI healthcare survey reveals a divided American public regarding the impact of artificial intelligence on patient well-being. While many Americans are open to AI handling administrative tasks, a significant wariness emerges when the technology moves into more critical medical decisions. The Pew survey AI healthcare public opinion indicates that while tasks like scheduling appointments, ordering prescription refills, and even translating for patients receive broad acceptance, this comfort level is contingent on human oversight. For instance, approximately eight out of ten adults are receptive to AI-powered translation, with a majority preferring a human to review the output. Similarly, the public largely supports AI analyzing medical scans or explaining laboratory results, provided a human professional verifies the AI's work.

This nuanced perspective highlights a general willingness to embrace AI as a supportive tool, particularly for reducing paperwork and improving efficiency. However, the American public draws a clear line when it comes to AI operating autonomously in sensitive medical contexts. The findings underscore a prevailing desire for human involvement to ensure accuracy and accountability, suggesting that the integration of AI into healthcare will need to carefully navigate these public expectations to foster trust and adoption.

Public enthusiasm for artificial intelligence in healthcare diminishes sharply when the technology is proposed for high-stakes decisions, such as determining a cancer diagnosis or adjudicating insurance claims. A substantial 41% of respondents expressed a desire for AI to have no role whatsoever in diagnostic processes, and 46% rejected any AI involvement in deciding insurance claims. Critically, only a mere 3% to 5% of individuals would feel comfortable with a machine making these critical decisions independently. This strong preference for human judgment extends to the doctor-patient relationship AI impact, with a clear majority of 60% reporting discomfort if their physician were to rely on AI for diagnoses and treatment recommendations, compared to only 30% who would feel at ease.

This widespread apprehension about AI's role in critical medical judgments also extends to concerns about data security. Over half of adults, 51%, believe that AI will weaken the security of personal health information, with a stark contrast to the mere 8% who anticipate an improvement in privacy. These sentiments carry significant AI in healthcare regulatory implications US, suggesting that any widespread deployment of AI in these sensitive areas without robust human oversight and stringent data protection measures could face considerable public backlash and potential legal challenges. The survey clearly indicates that trust in human professionals remains paramount for the American public in healthcare.

The Pew survey AI healthcare public opinion also reveals a complex and evolving outlook on AI's overall effect on patient health, with the public almost evenly divided: 37% anticipate better outcomes, while 36% foresee worse ones, a split that has remained consistent since 2022. However, a notable shift has occurred among younger adults, specifically those under 30, who have grown considerably more pessimistic. Four years prior, 29% of this demographic expected AI to worsen health outcomes; that figure has now risen to 45%. Conversely, adults aged 65 and older have become more optimistic, with 42% now expecting improved results.

Beyond direct health outcomes, pessimism is pronounced regarding the human element of medicine, with approximately half of adults believing AI will negatively affect the patient-doctor relationship. While views on access are less bleak, with 31% expecting AI to increase service availability versus 20% anticipating the opposite, opinions on costs and diagnostic accuracy remain mixed, with many respondents expressing uncertainty. Furthermore, public confidence in institutions outside of direct medical care is notably low. Majorities expressed little to no trust that health insurers, the federal government, AI companies, or even state governments would ensure AI is utilized in patients' best interests. Doctors, however, fared better, with 26% expressing a great deal or quite a bit of confidence, and another 34% reporting some confidence. These findings are not uniform across demographics; Asian Americans, individuals with higher incomes, and those with more education consistently show greater openness to AI and optimism about its effects, whereas lower-income adults and women tend to be more cautious.

This Pew survey provides critical insight into public sentiment regarding AI in healthcare, which lawyers advising healthcare providers, AI developers, and insurers should consider for risk assessment and compliance strategy. The strong public preference for human oversight in high-stakes medical decisions (diagnosis, insurance claims) indicates potential areas for future litigation or regulatory focus, necessitating careful policy development and communication strategies for clients implementing AI.

Source: Original reporting via Courthouse News Service