Neural Notes: Google’s AI watermarking faces its first real test
Google’s AI watermarking system is under pressure. Here is what SynthID’s limits mean for businesses, compliance and trust in disclosure.
Welcome back to Neural Notes, a weekly column where I look at how AI is influencing Australia. In this edition: a developer may have found a fatal flaw in Google DeepMind’s SynthID AI watermark system. Here’s why that’s a big deal for businesses.
A developer has claimed to have reverse-engineered Google DeepMind’s SynthID watermarking system.
Sign up to continue
You've reached your article limit. Create a free account, or log in to continue.
Get our free newsletters in your inbox and extended article access by submitting your email.
Comments