Bradley Merrill Thompson, Member of the Firm in the Health Care & Life Sciences practice, in the firm’s Washington, DC, office, was quoted in POLITICO Morning eHealth, in “New AI White Paper,” by Darius Tahir.
Following is an excerpt:
The agency is trying to figure out how to regulate artificial intelligence in medicine. In a new white paper Tuesday, the agency notes that advanced AI might be able to update itself in the future.
While it’s a potentially powerful approach — the agency contemplates software that gradually trains itself to better spot cancerous skin lesions — it’s also a tricky one. Traditionally the agency approves a product and substantial revisions require additional clearance. Software updating itself poses a tricky problem, and the agency is soliciting views about how it should scrutinize such products. (One possibility: The agency is thinking about requiring a “predetermined change control plans,” which specifies how the product might evolve over time.) …
Not everyone was wholly pleased. Epstein Becker Green lawyer Bradley Merrill Thompson generally agreed it was positive that the agency was considering the subject at all. But he worried that the agency was “becoming really good at coming up with ideas and not so good about carrying through with them.”
In particular, he flagged inconsistencies — whether the process would require draft guidance or new legislative authority — in the agency’s materials that suggested its strategy wasn’t entirely clear.
