T8-AT-012MEDIUM

Synthetic Media Support

T8 · External Deception & Misinformation →
Risk score185
RatingMedium
Procedures10
Severity
Mechanism

Synthetic media support is the dedicated text-generation layer that legitimizes fabricated images, audio, and video produced elsewhere (notably via T9). It manufactures the validating articles, backstories, corroboration, "expert" analyses, fake provenance, plausible metadata narratives, verifying fact-checks, timelines, and witness corroboration that convert a raw synthetic asset into a believed artifact. It works because audiences seldom assess media in isolation; they weigh surrounding explanation and apparent verification.

Detection
  • Media forensics on the underlying asset: Run synthetic-media detection on the image/audio/video the text is defending
  • C2PA / Content Credentials verification: Authenticate real provenance rather than trusting narrated provenance; fabricated provenance claims lack valid manifests
  • Metadata-claim verification: Independently inspect actual file metadata rather than accepting a narrative description of it
  • Fact-check authenticity verification: Confirm any validating "fact-check" comes from a recognized verifier
Mitigation
Provenance/capture credentials (C2PA) for genuine mediaHIGH
Synthetic-media detection at ingestMEDIUM
Verified-fact-check labelingMEDIUM
Coordinated-amplification detectionHIGH
Chaining

This technique is the natural partner of T9 synthetic media and deepfakes — it is invoked specifically to wrap T9 output in legitimacy, and it overlaps heavily with deepfake narrative creation (T8-AT-004). It draws on synthetic evidence (T8-AT-002) for documentary support, authority impersonation (T8-AT-001) for "official" validation, and synthetic testimony (T8-AT-008) for witness corroboration.

Framework mapping
OWASP LLMLLM09
MITRE ATLASAML.T0048
Open in the technique browser →