In a timely publication just ahead of the summer holiday season, the Oslo Law Review has released a research article titled "Doctor Chatbot: The EU's Regulatory Prescription for Generative Medical AI." This insightful piece, authored by Mathias Hauglid, a doctoral research fellow at UiT and legal advisor in SPKI, and Tobias Mahler, a professor at UiO, sheds light on the regulatory aspects surrounding the e.g. use of large language models and chatbots in medical applications.
The article draws attention to the potential regulatory gaps that exist between the AI Act and the Medical Device Regulation, highlighting the risks associated with medical AI systems falling through these cracks. Two case studies are examined to illustrate varying degrees of 'generativity' in medical AI systems, with the renowned Doctor Chatbot serving as one of the case studies.
The key topics explored in this article include:
- The interplay and interaction between the AI Act and the Medical Device Regulation.
- The specific risks posed by generative medical AI systems, particularly those employing Large Language Models.
- The classification of generative medical AI systems in terms of risk, according to the AI Act and the Medical Device Regulation.
- The considerations involved in determining whether an AI system should be classified as a medical device.
To delve deeper into this important subject, read the full article here. This comprehensive study provides valuable insights into the regulatory challenges faced by the implementation of large language models and chatbots in the field of medicine.