FDA is actively monitoring the use of AI and machine learning (ML) software in medical devices and clinical developments, and has taken a number of first steps in building out a regulatory framework. In 2021, FDA published the AI and Machine Learning Software as a Medical Device Action Plan, which outlines FDA’s approach to regulating AI/ML technologies and next steps in furthering oversight, as well as the Good Machine Learning Practice for Medical Device Development: Guiding Principles, to inform the development of Good Machine Learning Practice. FDA also released a database of AI/ML-enabled medical devices currently marketed in the United States and will update this list as FDA reviews and authorizes additional AI/ML-enabled medical devices going forward. While AI technology continues to advance far ahead of a regulatory framework, FDA has been collaborating with international counterparts and actively seeking industry and consumer feedback in developing its approach to oversight. Going forward, we can expect to see additional guidance from FDA, with a continuing emphasis on transparency and diversity as well as ongoing requests for industry feedback. FDA intends to publish draft guidance on Marketing Submission Recommendations for a Change Control Plan for AI/ML-Enabled Device Software Functions in 2022, although this guidance has been pushed to the B-list in terms of priority.
This post is a part of a series on trends in the artificial intelligence space for 2022, authored by MoFo lawyers.