Meta's deepfake moderation isn't good enough, says Oversight Board
Posts from this topic will be added to your daily email digest and your homepage feed.
TL;DR
- Meta's AI content identification methods are not robust enough to handle rapid misinformation spread, especially during armed conflicts.
- The Meta Oversight Board is calling for an overhaul of how Meta labels and surfaces AI-generated content on its platforms.
- Current systems rely too heavily on self-disclosure and escalated review, failing to meet the demands of the online environment.
- The Board recommends improving misinformation rules, creating separate standards for AI content, developing better detection tools, and increasing AI content labeling.
- Emphasis is placed on scaling AI content labeling, including 'High-Risk AI' labels and improving C2PA adoption for user transparency.
- Meta is reportedly inconsistent in implementing C2PA, even for content generated by its own AI tools.