Patients prefer healthcare providers' AI agents to public chatbots, with human oversight non-negotiable, survey finds

A recent Salesforce survey, encompassing over 3,200 patients across eight countries, reveals a notable shift in patient attitudes towards artificial intelligence in healthcare. The findings show that 61% of global patients are comfortable with AI in healthcare contexts, and 64% would be willing to share their full medical history with AI for quicker diagnoses. However, this growing acceptance comes with significant caveats regarding trust and governance. Patients are three times more likely to trust AI agents that are integrated into their healthcare provider's secure portal compared to general public chatbots. This preference underscores the importance of institutional accountability and the context provided by a trusted healthcare provider. Patients are open to AI handling logistical tasks such as billing and rescheduling, with almost half preferring AI for these functions to avoid wait times, particularly outside of regular office hours. Despite the willingness to engage with AI for convenience, human oversight remains critical. A substantial majority—89% of patients for administrative support and 90% for medical support—emphasize the necessity of a clear "escalate to human" option. Furthermore, 91% believe they should retain the right to opt out of AI-driven clinical recommendations. The primary concerns for patients regarding AI in healthcare are the accuracy of diagnosis or treatment (36%) and the privacy and security of their health data (30%). The survey highlights that while AI offers immense potential to streamline healthcare operations and improve access, its successful implementation must prioritize patient trust through robust governance, data security, and the assurance of human intervention.