Google’s SynthID scare shows AI watermarking is still not a trust layer

Google image used for GCATS article coverage.
Google

The latest SynthID questions reinforce that provenance tooling is useful, but not strong enough to carry trust on its own.

Join the conversation