Public perceptions of health information generated by AI: A research study

Abstract

Artificial Intelligence (AI) integration in clinical practice has intensified in the last few years, from systems analysing and interpreting existing data to generative AI systems capable of creating new information and offering new possibilities for patient communication.1 However, the public’s perception of AIgenerated health information remains largely unexplored. This study aimed to assess public trust in AI-generated health information, identifying influencing factors on their trust and evaluating the accuracy of AI-produced content. A mixed-method approach was employed, involving a survey distributed via social media to individuals with recent access to health information. Results revealed that while the public knew AI systems’ capabilities, their trust in AI-generated content was moderate. Key concerns included: the accuracy of the information, potential biases in AI algorithms, and ethical issues related to privacy. Results showed that transparency, healthcare professional endorsements, and clear evidence of accuracy are critical in building trust in AI-generated health information. Addressing these concerns is essential for successfully integrating AI into patient communication, to enable the reliability and use of AI as an ethical tool in healthcare.