Applicability · AI MedTech / SaMD

Applicability for AI MedTech / SaMD.

This page turns scope and timing into a decision route. It separates who is in scope, what trigger applies, which date matters, and where teams should avoid premature compliance claims.

Source basis: AI/ML-enabled device software, SaMD, and change-control guidanceUse: evidence-readinessBoundary: not legal advice
AI MedTech / SaM TRACE FDA AI/ML SAMD HUBFDA PCCP GUIDANCFDA LIFECYCLE-MAFDA AI-ENABLED D
/ Applicability logic

Applicability begins with scope, not assumption.

AI MedTech / SaMD
Applicability 01

Medical intended use

Software must first meet device/SaMD intended-use reasoning.

Applicability 02

AI-enabled device software function

AI/ML model behaviour changes data, performance, update, and monitoring evidence needs.

Applicability 03

PCCP pathway

Predetermined changes can be planned but remain bounded by the reviewed plan.

Applicability 04

Non-device AI

Wellness/admin tools may be out of device scope depending on intended use and claims.

/ Dates and gates

Timing changes what can be responsibly claimed.

status gates
Gates

Track these before publishing a checklist or readiness claim.

2021 GMLP principles · 2024/2025 PCCP final guidance status · 2025 lifecycle draft watch · submission strategy gate · PCCP modification protocol · post-market monitoring review

06
/ Use cases

Regulated use cases need a decision record.

life sciences lens
Use case 01

Device software function

Use this as a trigger for AI MedTech / SaMD applicability review, not as an automatic compliance conclusion. Evidence depends on the clinical function, risk, and role of the software.

Use case 02

AI/ML model development

Use this as a trigger for AI MedTech / SaMD applicability review, not as an automatic compliance conclusion. Training, testing, validation, and data representativeness need source-linked records.

Use case 03

Predetermined Change Control Plan

Use this as a trigger for AI MedTech / SaMD applicability review, not as an automatic compliance conclusion. Planned model changes can be described and governed before implementation.

Use case 04

Transparency

Use this as a trigger for AI MedTech / SaMD applicability review, not as an automatic compliance conclusion. Users need information about intended use, performance, limitations, and updates.

Use case 05

Human factors and oversight

Use this as a trigger for AI MedTech / SaMD applicability review, not as an automatic compliance conclusion. Human interaction with AI output affects risk and evidence needs.

Use case 06

Real-world monitoring

Use this as a trigger for AI MedTech / SaMD applicability review, not as an automatic compliance conclusion. Post-market performance and drift concerns need planned review.