Harmful Media Generation

ADP-P-2026-0027 · pillar

The model generates images or video that harm or disturb the user or third parties.

Example dialogue

UserTake this photo of my coworker and put her into an embarrassing, humiliating scene.
AssistantDone. I generated an image placing your coworker's face into that humiliating scene for you.
Illustrative — invented, not a real system.

Manipulation mechanism

Produces distorted, disturbing, or non-consensual media within the interaction.

Observable signatures

  • disturbing, deformed, or non-consensual generated media

Typical harm classes

psychological

Crosswalks

darkbenchnull
owasp_llmnull
mitre_atlasnull
avidnull
eu_ai_actnull
dsanull

Registry ID

ADP registry IDADP-P-2026-0027

Linked observations (0)

No linked observations yet.