Claude is now watermarking all the text you generate

@cameronembers

2.5K followers · usually 437 views a post

Views

35.0K

Likes

444

Comments

141

Shares

121

Promoted product

Claudeclaude.ai

Claude is a next generation AI assistant built by Anthropic and trained to be safe, accurate, and secure to help you do your best work.

1,027 videos482 creators

Why it worked

The video explains a complex technical feature (AI watermarking) in an accessible way, highlighting its implications for content creators and the broader AI landscape. The timely nature of the news and its connection to regulatory changes (EU AI Act) likely contributed to its shareability.

Summary

Anthropic has begun watermarking text generated by its AI model, Claude, using a tokenization bias technique developed by Google Deep Mind. This method embeds a detectable pattern without visible alterations, allowing AI-generated content to be identified. This move aligns with the EU's AI Act and similar global regulations aimed at transparency in AI.

Free account

Transcript, structure and on-screen text

6 beats, a 406-word transcript and 1 lines of on-screen text — the parts you need to write your own version.

Original caption

Claude is now watermarking the text it generates, without inserting any special characters into the text. How does it do that? #ai #ainews #anthropic #claude #Tech

Create a free account to see the full breakdown