AIBleepingComputerabout 2 hours ago
How Anthropic plans to watermark Claude's AI-generated text
1 min read
Anthropic will watermark Claude-generated text worldwide to comply with the EU AI Act, using Google DeepMind's SynthID-Text approach. The invisible watermark alters randomness during token selection, leaving a detectable pattern. A detection API will estimate Claude's involvement, with no impact on quality, speed, or cost.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyAnthropic
Summary by ByteBrief
