SynthID and C2PA solve different problems
Google DeepMind describes SynthID as an imperceptible watermark embedded into generated or altered images, audio, text, and video, with detection handled by SynthID technology. C2PA instead stores signed provenance assertions that compatible validators can read.
An asset can have one, both, or neither. Removing metadata does not necessarily remove a robust pixel-level watermark; conversely, reading metadata does not reveal a private watermark signal.
How to check responsibly
AI Origin Check can inspect C2PA and ordinary metadata locally. It cannot independently verify a private SynthID signal. Where available, upload the content to Google’s official Gemini or SynthID verification flow and ask whether the signal is detected.
- Use Google’s official detector for a SynthID claim.
- Use the local file checker for C2PA and editable metadata.
- Treat any single negative result cautiously after heavy editing or conversion.
Primary sources and further reading
For technical and policy detail, consult the C2PA specification, Content Authenticity Initiative developer documentation, Anthropic transparency material, Google DeepMind’s SynthID overview, and the European Commission’s AI transparency guidance.