AI Assurance Lab Concept Leaves Potential Regulatory Gap

In an interview with Medtech Insight sister publication the Pink Sheet, Troy Tazbaz, director of the US FDA's Digital Health Center of Excellence, says the agency lacks authority to regulate assurance labs, which would be used to help AI developers ensure their models are working correctly before submitting them for approval.

AI Privacy Health Data
The FDA is investing in AI lifecycle management, Tazbaz said. • Source: Shutterstock

Assurance labs, a potentially key part of the development of artificial intelligence for medtech industy use, are not part of the official regulatory process yet, which could raise questions for the US Food and Drug Administration as the technology pushes further into the industry.

Key Takeaways
  • The FDA’s AI regulatory process currently cannot oversee assurance labs that would help test the algorithms to ensure they are working as intended.

AI models in development for medical product research and clinical practice are expected to need oversight before and after deployment...

Read the full article – start your free trial today!

Join thousands of industry professionals who rely on Medtech Insight for daily insights

  • Start your 7-day free trial
  • Explore trusted news, analysis, and insights
  • Access comprehensive global coverage
  • Enjoy instant access – no credit card required

More from AI

From Risk Aversion to Real-World Impact: How Medtech Firms Can Harness AI

 
• By 

Medical device companies are increasingly using AI to improve regulatory affairs strategies but face barriers such as risk aversion, regulatory clarity and data governance issues. Experts suggest a design control approach and human oversight to ensure successful AI integration.

Medtech Sees Bigger Deals As Early Funding Tightens; LSI Is ‘Cautiously Optimistic’ For 2026

 
• By 

Medtech funding in 2025 sees a trend of fewer, yet larger deals with total funding of $3.6bn in Q1, $2.6bn in Q2 and $2.9bn in Q3, said LSI’s Nick Talamantes. In 2026, he expects a continued investor focus on more mature firms in areas of oncology, cardiology and neurology.

Getting Comfortable With The Black Box: Explainable AI Techniques Build Trust

 

A subfield of AI aims to explain why complex models make predictions, which will increase their interpretability and auditing ability, AI researcher Su-In Lee said during a National Academies of Sciences workshop.

FDA Staff Warn Device AI Hallucinations Could Harm Patient Care

 

Hallucinations are a common problem in AI and pose significant challenges to medical devices powered by the technology. A paper authored by several FDA staffers argues that focusing on how hallucinations impact patients is the best approach.

More from Digital Technologies

FDA Staff Warn Device AI Hallucinations Could Harm Patient Care

 

Hallucinations are a common problem in AI and pose significant challenges to medical devices powered by the technology. A paper authored by several FDA staffers argues that focusing on how hallucinations impact patients is the best approach.

Masimo Awarded $634M In Apple Patient Monitoring Patent Fight

 
• By 

A California jury found in favor of device firm Masimo in a long-running legal battle against the tech giant. Apple has already announced plans to appeal.

Digital Health Roundup: BCIs Surge Ahead, Oura Eyes BP Clearance, AI Faces New Scrutiny

Brain-computer interfaces advance toward trials and commercialization, Oura pushes for FDA-cleared blood pressure monitoring, and regulators weigh AI’s expanding role in mental health and diagnostics amid rising safety concerns.