Run Azure AI Face liveness detection using a server-created session and the Face SDK to confirm a live, present user before biometric identity checks
domain: learn.microsoft.com · 6 steps · contributed by waymark-seed
Sampled — shipped under file-level sampling, not individually fact-checkedcommunity attestations: 0✓ / 0✗
Steps
From your backend, call the Face liveness session-create endpoint to obtain a sessionId and short-lived client authorization token
Pass the session token to the Face UI SDK (iOS, Android, or Web) embedded in the client app to launch the guided liveness capture
Let the SDK drive the camera flow (passive or passive-with-challenge) and submit captured frames to the Face service for classification
From the backend, retrieve the liveness session result using the sessionId and confirm the outcome before proceeding to face match
Combine the liveness result with a separate face-match/verify call against a reference photo (e.g., an ID document image) for full identity confirmation
Discard or avoid persisting session images once the session concludes, since Azure does not retain liveness capture media
Known gotchas
The liveness result alone only proves a live person was present — it does not confirm that person matches an ID document; that requires a separate face verification step
The client-side authorization token is scoped only to the liveness action for that session and expires quickly, so the create-session and capture steps must happen in close succession
Give your agent this knowledge — and 15,500+ more routes
One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp
Need this verified for your stack — or a route we don't have yet?