AdvaMed: AI medtech companies facing more risk with uncertainties at FDA

Medtech companies developing artificial intelligence (AI) products are taking on more risk as the Trump administration’s actions at the US Food and Drug Administration (FDA) have halted new AI-related guidances from being published, according to the lobby group AdvaMed.

Over the past several months, the Trump Administration has taken steps that have frozen guidance development at FDA, including issuing an executive order requiring 10 regulations, rules, or guidances be withdrawn before a new one can be published. The administration also recently fired staff who were critical to developing new guidances as part of its reduction-in-force (RIF) initiatives. (RELATED: Thousands of FDA staff fired in latest RIF, Regulatory Focus 1 April 2025 and Trump's 10-for-1 order puts pressure on FDA to find regulations to nix, Regulatory Focus 7 February 2025)

With the uncertainty surrounding FDA’s thinking, medtech companies are taking on more risks in their AI-enabled product development, according to Robert Cohen, president of Stryker’s digital, robotics, and enabling technology division. During a virtual press conference on 22 April to discuss a new AdvaMed AI policy roadmap, he noted that medtech companies have not slowed down their work but are making different risk calculations in response to the uncertainties created by the administration’s actions.

“The strategic plans we all have within our companies, and the project priorities, and the roadmaps, they remain unchanged,” said Cohen, who is on the board of AdvaMed’s Digital Health Tech Division.

Cohen said that while companies are developing their products, they are cognizant that FDA could develop policies in the near future that could derail their plans. He said companies need to ask themselves if they are flexible enough to weather any such policies.

“It's not necessarily change, but we are sitting here with some, healthy anxiety that we expect things to come,” said Cohen.

Cohen noted FDA has already granted marketing authorization to hundreds of AI-enabled devices and applications. He noted that there are plenty of predicate products on which companies can base their decisions and assess risks.

Cohen said companies trying to minimize their risk, especially smaller companies, should focus on where they are getting their data, how they plan to validate their model, and prepare themselves to answer every potential question the FDA might have for them. If the agency takes a more conservative approach to AI, answering questions such as where the data came from and how it was used to train the AI model becomes even more important, he added.

Nanowear CEO Venk Varadan, who also serves on the Digital Health Tech Division board, agreed with the recommendations and said startups need to look at what has come before and learn from the challenges that companies such as his have gone through to bring AI devices to market. He noted that companies often hide their protocols and other information to protect themselves from their competitors, but they need to come together as a community to try help each other out as much as possible.

“We do all share our best practices from small company to large company from medtech company to tech company because US sort of sharing that knowledge is how we're going to get more technologies into the market small and big,” said Varadan.

Members of AdvaMed’s Digital Health Tech Division discussed an AI policy roadmap during the press conference that they have presented to congressional lawmakers. It includes recommendations to allow for more data protection flexibility, especially regarding Health Insurance Portability and Accountability Act (HIPAA) requirements. The group said it recommends updating HIPAA for the AI era to "Ensure that HIPAA standards allow for the sharing of the datasets needed to train, test, validate, and re-train AI models while preserving patient privacy. The current HIPAA de-identification methods (authorization, safe harbor, and expert determination) stifle the high-volume data usage and sharing that can optimize the development of safe and accurate AI models."

The roadmap also includes recommendations for FDA, such as implementing the agency's existing predetermined change control plan (PCCP) authority given to it under the 2022 Food and Drug Omnibus Reform Act (FDORA), issuing AI guidances in a timely manner, and ensuring that its policies are harmonized with other global regulators.

"With the new administration, a new Congress coming in, we thought it was really important to help them lay out a roadmap for how we take a look at AI, future regulations and legislation,” said Shaye Mandle, executive director of the Digital Health Tech division. “We wanted to cut through some of the noise and be able to put something in front of policymakers to get a better sense where medtech and big tech are coming together, where those primary issues are."

Varadan echoed the sentiments and said the roadmap is an opportunity to engage the new administration as it is changing the way FDA operates.

“The new administration is seismically shifting, let's call a spade a spade, but at some point, there will be a steady state on how they're structuring their institutions going forward,” said Varadan. “We recognize that the traditional sort of FDA person might look different in a year, but we have to continue being glass half full and optimistic on bringing these technologies to our citizenry, and that requires working with a traditional looking FDA as well as the new looking FDA and so we have to be agnostic on that.”

AI policy roadmap

AdvaMed: AI medtech companies facing more risk with uncertainties at FDA

Related topics

Recommended content

Welcome to the new RAPS Digital Experience