#watermarking
2 stories
Google opens SynthID detector to everyone
Google's tool for spotting AI-generated content now checks watermarks from OpenAI, Nvidia, Kakao and Apple on one website.
Ars Technica · 1m
AI watermarks may weaken safety guardrails
New research reveals that watermarking techniques designed to identify AI-generated text can paradoxically make models more vulnerable to harmful prompts.
Ars Technica · 2m