---
title: "Call Deepfake Detection Result"
method: POST
path: "callDeepfakeDetectionResult"
tags: ["Callbacks"]
---

# Call Deepfake Detection Result

`POST callDeepfakeDetectionResult` (webhook)

This webhook uses Telnyx headers (telnyx-timestamp, telnyx-signature-ed25519) that are compatible with Standard Webhooks specification for SDK generation. Custom validation logic can map these to standard webhook-timestamp and webhook-signature equivalents. See https://github.com/standard-webhooks/standard-webhooks for details.

## Headers

- `telnyx-timestamp` integer, required — Unix timestamp for replay attack prevention
- `telnyx-signature-ed25519` string, required — Ed25519 signature of timestamp|payload for verification

## Payload

- CallDeepfakeDetectionResultEvent
  - `data` CallDeepfakeDetectionResult
    - `record_type` 'event' — Identifies the type of the resource.
    - `event_type` 'call.deepfake_detection.result' — The type of event being delivered.
    - `id` string, uuid — Identifies the type of resource.
    - `occurred_at` string, date-time — ISO 8601 datetime of when the event occurred.
    - `payload` object
      - `call_control_id` string — Call ID used to issue commands via Call Control API.
      - `connection_id` string — Call Control App ID (formerly Telnyx connection ID) used in the call.
      - `call_leg_id` string, uuid — ID that is unique to the call and can be used to correlate webhook events.
      - `call_session_id` string, uuid — ID that is unique to the call session and can be used to correlate webhook events. Call session is a group of related call legs that logically belong to the same phone call, e.g. an inbound and outbound leg of a transferred call.
      - `client_state` string, nullable — State received from a command.
      - `result` 'real' | 'fake' | 'silence_timeout' — Detection outcome. 'real' = human voice, 'fake' = AI-generated voice, 'silence_timeout' = no analyzable speech detected before timeout.
      - `score` number, float, nullable — Probability that the audio is AI-generated, from 0.0 (likely real) to 1.0 (likely deepfake). Based on the model's aggregated confidence across analyzed audio frames. Null for silence_timeout.
      - `consistency` number, float, nullable — Percentage (0-100) indicating how consistently the model classified the audio across frames. High consistency (>90%) means confident classification throughout; low consistency suggests mixed signals. Null for silence_timeout.

## Acknowledgement `200`

The server must return an HTTP 200, otherwise we will reattempt delivery at the failover webhook URL.

---

[API](https://skmtc.net/team-telnyx/apis/telnyx-api-2.md) · [All operations](https://skmtc.net/team-telnyx/apis/telnyx-api-2/llms.txt) · [OpenAPI document](https://skmtc-service-staging.skmtc.workers.dev/v1/apis/team-telnyx/telnyx-api-2/versions/8f5f4e537994/schema)