|
Work item:
|
X.sem-ddm
|
|
Subject/title:
|
Security evaluation methodology for DeepFake detection model
|
|
Status:
|
Under study
|
|
Approval process:
|
TAP
|
|
Type of work item:
|
Recommendation
|
|
Version:
|
New
|
|
Equivalent number:
|
-
|
|
Timing:
|
2028-Q1 (Medium priority)
|
|
Liaison:
|
ITU-T SG21, JVET , ISO/IEC JTC 1/SC 42
|
|
Supporting members:
|
CAICT, China Telecom, China Unicom
|
|
Summary:
|
This Recommendation specifies a security evaluation methodology for deepfake detection models used in image, video, and audio authenticity assessment.
This Recommendation defines security threats, evaluation objectives, metrics, test scenarios, and assessment procedures for evaluating deepfake detection models. It addresses security risks including adversarial manipulation, data poisoning, model extraction, backdoor attacks, and other threats relevant to deepfake detection systems.
The methodology considers the domain-specific characteristics of deepfake detection, including multimodal content analysis and the continuous evolution of generative technologies. It is intended for use by developers, evaluators, platform operators, service providers, and regulatory authorities to support the assessment and deployment of trustworthy deepfake detection systems.
|
|
Comment:
|
-
|
|
Reference(s):
|
|
|
Historic references:
|
|
Contact(s):
|
|
| ITU-T A.5 justification(s): |
|
|
|
|
First registration in the WP:
2026-06-10 08:28:36
|
|
Last update:
2026-06-18 14:01:40
|
|
