AI in Regulatory Science: Conference Report
The 9/18-9/19/2024 Global Summit on Regulatory Science was hosted by FDA with the theme “Is Regulatory Science Ready for Artificial Intelligence?” A report from the conference, published in the open-source NPJ Digital Medicine, includes 10 key principles for regulating artificial intelligence (AI) described by former FDA commissioner Robert Califf. His principles address: global and interagency collaboration, flexibility in regulation, regulatory boundaries, AI in medical product development, large language models and generative AI, lifecycle management, supply chain transparency, balancing innovation and scale, financial dynamics in healthcare, and local recurrent validation.
The paper also says that successful AI integration in regulatory science depends on overcoming key challenges related to data quality and availability, AI model validation, human-AI interaction, and accountability and responsibility.
“It is important to note that the challenges and opportunities for AI integration are dynamic and evolving, requiring flexible and adaptive strategies,” the report concludes. “A framework for maintaining human oversight, ensuring fairness and transparency, and continuously evaluating the impact of AI-assisted decisions is central to successful AI integration. More importantly, this should be a collaborative effort among global regulators to build a culture of responsible integration, and a commitment to the principles of good regulation, so that AI can be harnessed to meet 21st century demands in regulatory science.