Washington – September 18, 2026 -- The Association for Diagnostics & Laboratory Medicine (ADLM) has formally recommended that artificial intelligence tools used in clinical laboratories remain subject to the same validation, quality-system, and monitoring requirements that govern traditional laboratory testing under the Clinical Laboratory Improvement Amendments (CLIA). The recommendation was submitted in a comment letter responding to a federal request for information on potential updates to CLIA, the regulatory framework that has governed U.S. clinical laboratory testing since 1992.
ADLM urges CMS and CDC to close oversight gaps without duplicating FDA rules
The letter calls on the Centers for Medicare & Medicaid Services and the Centers for Disease Control and Prevention, the two agencies that administer CLIA, to establish federal oversight of AI-based laboratory tools that avoids overlap with the Food and Drug Administration. ADLM notes that the FDA may regulate specific software products or medical devices, while CLIA governs a laboratory's responsibility for delivering accurate and reliable test results.
Association warns AI models fail differently than conventional laboratory software
ADLM argues that traditional software errors typically affect every case meeting the same programmed conditions and can be tested against known expected outputs. AI-based models, by contrast, can produce case-specific errors that are harder to detect and troubleshoot. Generative AI tools introduce additional risk, according to the letter, including the potential to generate inaccurate or unsupported information, omit clinically relevant facts, or shift behavior after updates to the underlying model, prompt, or knowledge base.
Comment letter calls for risk-based CLIA updates rather than a separate AI regulatory track
ADLM's recommendations include distinguishing conventional software from AI models within CLIA and requiring laboratories to account for differences in validation and performance monitoring between the two. The letter also calls for a risk-based, technology-appropriate approach that sets clear quality expectations for AI tools while preserving laboratory directors' authority to determine scientifically appropriate validation methods. Where a facility independently analyzes patient-specific data or provides specialized interpretation that generates or contributes to a clinical test result, ADLM says those activities constitute part of the total testing process and should fall under CLIA oversight.
ADLM president says laboratory-based oversight remains essential as AI adoption grows
"AI has the potential to support tremendous advances in laboratory medicine, but innovation in this area must be balanced with the need to ensure test quality and patient safety," said ADLM President Dr. Stanley F. Lo. He added that laboratories already possess the expertise and quality systems needed to evaluate, implement, and continuously monitor AI tools.
ADLM's overall position is that AI-based tools should be assessed as part of the existing total testing process under CLIA rather than through a newly created, standalone regulatory structure.