SynthID makes Gemini a media provenance checker
Google’s SynthID lets Gemini inspect uploaded images, video, and audio for invisible watermarks and Content Credentials indicating whether Google AI was involved. A negative result does not prove that media is human-made or authentic.
The important shift is practical access, not a new detector: provenance checks are moving into everyday multimodal workflows, but they remain evidence rather than proof.
- –Invisible watermarks are designed to survive common edits such as compression, cropping, filters, and resizing
- –Content Credentials can provide richer origin and edit-history context than watermark detection alone
- –Gemini currently distinguishes Google AI content, so unsupported generators can still evade detection
- –Developers should treat “not detected” as inconclusive and combine provenance signals with source, metadata, and visual review
DISCOVERED
1h ago
2026-10-18
PUBLISHED
1h ago
2026-10-18
RELEVANCE
AUTHOR
Rob The AI Guy