Meta AI detector fails cropped images
Analysis based on 9 articles · First reported Apr 20, 2026 · Last updated Jul 10, 2026
The flaw may erode trust in Meta's AI content moderation capabilities, potentially affecting user engagement and advertiser confidence. However, the impact is limited as the tool is a preview and competitors face similar challenges.
Reuters analysis found that Meta's new AI detection tool, previewed alongside the Muse Image model, failed to identify 55% of its own AI-generated images after cropping. Meta acknowledged the limitation, noting the watermark may be lost with heavy cropping. The Oversight Board had previously called for stronger detection tools. Experts noted watermarking has limitations but is a step forward.
Set up alerts, explore entity relationships, search across thousands of events, and build custom intelligence feeds.
Open Dashboard