Meta made its own AI detection system. It should have just used Google’s
The Verge
Meta launched its own AI watermarking tool, Content Seal, but it offers fewer benefits and interoperability than existing alternatives like Google's SynthID.Summary
Meta recently introduced Content Seal, an invisible watermarking tool designed to flag images generated by its new Muse AI model. However, rather than developing a proprietary system that lags behind established alternatives, Meta could have adopted Google’s widely accepted SynthID, which is already used by competitors like OpenAI. Content Seal currently suffers from several limitations, including restricted platform integration, lack of video support at launch, daily usage limits, and poor interoperability with other detection standards like C2PA. Ultimately, the new system feels rushed and fails to provide unique benefits for improving AI transparency at scale.
(Source：The Verge)