Non-Consensual Synthetic Intimate Imagery
未經同意的合成親密影像
Operation & Monitoring
Risk Description
When a victim's public photographs are used to synthesize false content circulated in closed groups; due to unmitigated control gaps, by the time the victim becomes aware, the content has spread widely, triggering external stakeholder impacts and causing removal requests must be filed with each platform individually, and re-uploading cannot be prevented.
Framework Mappings
EU AI ActArt.50
NIST AI 600-1Obscene, Degrading, and/or Abusive Content
NIST AI RMFMANAGE 2.4
ISO/IEC 5338運作與監控
MIT AI Risk RepositoryDomain 4
Risk Treatment & Implementation Guidance
Deploy deepfake detection and generation-side blocking of synthetic intimate imagery; Establish platform governance and rapid-takedown collaboration to shorten spread-to-removal time; Provide victim remedy and legal-recourse support with records enabling cross-platform reporting