Nikhil Pradhan on FDA Draft AI Drug Guidance – 'The industry is very much evolving'

Foley & Lardner LLP senior counsel Nikhil Pradhan shared insight on recent draft guidance from the U.S. Food and Drug Administration in the MedCentral article, “FDA Draft Guidance Addresses Drug Submissions that Use AI Data.”

Pradhan described the FDA’s guidance as overall striking a good balance between promoting innovation and ensuring AI model credibility. To comply with the framework if it is finalized, Pradhan recommended that drug developers rely on their historical experience providing data during the approval process to inform how they will meet its requirements with potential disclosure of valuable information.

“If there are uncertainties about comfort level with disclosure, it may be useful to reach out to the engagement groups the FDA has identified at the end of the guidance for specific subject matter areas,” he suggested.

Pradhan said AI models can be used during drug development to generate reports, surface patterns, or analyses to meet the FDA’s data requirements. He noted how the use of AI is continuing to expand in ways that “include extracting data from the drug development process that is relevant to the agency’s criteria, generating synthetic data that can be used to test AI models, and using AI to correct for AI-related challenges such as data drift or bias.”

“The industry is very much evolving,” Pradhan added. “Some companies with lab automation techniques are positioning themselves as ‘data as a service’ to support therapy developers; other companies may seek to pursue end-to-end solutions rather than focusing on one step in the pipeline.”