Anthropic defends and explains its AI watermarking feature
By
Stephen Council
You're currently following this author!
Want to unfollow? Unsubscribe via the link in your email.
Anthropic defends and explains its AI watermarking feature
Anthropic addresses concerns about its new watermark, which will affect Claude users globally. It said the watermark is to comply with the EU AI Act.
TL;DRAI
Anthropic watermarked Claude text to comply with the EU AI Act via imperceptible word-choice patterns, applied globally. The watermark identifies only Claude involvement—not AI generally—and weakens on code, raising governance questions for enterprises.
892 words~4 min read
