Decision Tree for FDA SaMD Compliance

Follow the steps below to determine if your AI/ML app requires FDA Software as a Medical Device (SaMD) compliance.

Does the app perform a medical function (diagnose, treat, cure, mitigate, or prevent disease)?

Summary

If the app performs a medical function and assists in clinical decision-making, FDA SaMD compliance applies.

If the app uses AI/ML for real-time decision-making, especially with adaptive learning, FDA’s AI/ML regulatory framework must be followed.

If the app is for wellness, education, or general health tracking, SaMD does not apply, but FTC, HIPAA, and cybersecurity regulations may be relevant.

Understanding FDA SaMD Compliance for AI/ML Applications

The FDA Software as a Medical Device (SaMD) Decision Tree helps developers determine whether their AI/ML-powered applications require FDA compliance. SaMD regulations apply to software that plays a direct role in diagnosing, treating, curing, mitigating, or preventing diseases.

If your AI/ML model is continuously learning and adapting, it may require a Predetermined Change Control Plan (PCCP) under FDA guidelines. Apps used strictly for wellness, education, or general health tracking may not fall under SaMD but should comply with other regulatory frameworks, such as FTC, HIPAA, and cybersecurity standards.

By following this structured decision tree, you can better understand whether your software product falls under FDA oversight and what steps may be necessary to ensure regulatory compliance.