.png)
.svg)
In early January 2026, the FDA issued revised guidance significantly expanding exemptions for certain digital health products from active regulatory oversight. The update reflects the agency’s continued risk-based approach, focusing regulatory resources on software functions that present higher patient safety risk while reducing burden for low-risk digital health technologies.
Wearables: More wearable devices measuring physiological parameters (e.g., blood pressure, blood glucose) are now considered "general wellness" products rather than medical devices, provided they do not make medical claims.
Clinical Decision Support (CDS): Updated guidance allows AI-enabled software that provides single treatment recommendations to enter the market without full FDA review in certain contexts.
Impact on Pre-Market Submissions
Under the revised guidance, a broader range of digital health software -particularly wellness, administrative, and certain clinical decision-support functionalities - may no longer require traditional marketing submissions such as 510(k)s. However, manufacturers must still carefully assess intended use, claims, and software functionality to confirm eligibility for exemption. AI-enabled products with diagnostic, predictive, or treatment-driving outputs remain subject to FDA review, with clear expectations for transparency, validation, and lifecycle controls.
Implications for Manufacturers
For manufacturers, the expanded exemptions offer faster market entry, reduced regulatory costs, and increased flexibility in product iteration. At the same time, they place greater responsibility on companies to self-classify accurately, maintain robust quality systems, and ensure claims remain within exempt boundaries. Overstating AI capabilities or clinical impact remains a significant regulatory risk.
Challenges and Regulatory Pitfalls
Manufacturers implementing AI-enabled devices continue to face challenges related to data quality, bias, model updates, and post-market performance monitoring. Regulators, meanwhile, must contend with limited visibility into exempt products, difficulties evaluating adaptive algorithms, and the risk that rapidly evolving AI software may drift into higher-risk use cases without timely oversight.
For more information, visit https://eemedicals.com/
.svg)
.svg)
.svg)