Can readers spot the AI impostor in healthcare writing?

The use of artificial intelligence (AI) writing assistants in the healthcare industry is becoming increasingly prevalent. These tools can help medical writers to generate content more quickly and efficiently, but they also raise concerns about the accuracy and completeness of the information that is produced. This study investigated whether readers can distinguish between health-related texts written by humans and those generated by AI writing assistants. A survey of 164 respondents found that slightly more than half could correctly identify the source of the healthcare text. Differences between healthcare professionals and non-healthcare professionals were not statistically significant. Medical writers were better at recognising that a text had been written by an AI model than were non-medical writers (P<.05). These findings suggest that it is important for organisations to establish clear guidelines regarding the use of AI writing assistants in healthcare. The authors of health-related content should be required to identify whether their work has been completed by a human or an AI writer, and organisations should develop processes for evaluating the accuracy and completeness of AI-generated content. This study has several limitations, including the small sample size. However, the findings provide valuable insights into the need for organisations to develop clear guidelines for their use.