AI News

    Anthropic explains how Claude’s invisible text watermarks will work

    Anthropic adopts DeepMind's SynthID to embed invisible watermarks into AI-generated text for better transparency.

    Anthropic has clarified how it's planning to apply invisible watermarks to Claude-generated text in order to comply with Europe's AI transparency rules. On Friday, Anthropic announced that Claude's text marking system is "a version of the SynthID-Text approach" - an open-source w...

    TVA

    The Verge AI

    Source

    The Verge AI
    Anthropic explains how Claude’s invisible text watermarks will work
    AI Analysis

    Generated insight by OnAirToday AI

    Anthropic's adoption of Google DeepMind's SynthID marks a critical step toward standardizing provenance in generative AI content. By embedding statistical patterns directly into word probability, the industry is moving closer to reliable methods for distinguishing human authorship from machine-generated text.

    Powered by AI • Not a substitute for human analysisMore Insights →
    Read the full story at The Verge AI →
    Comments
    Share
    Filed under:AI News
    #Tech Policy#Content Authenticity#Ai Governance

    Comments

    Keep comments relevant and respectful. No links, advertising, self-promotion, or spam.

    0/1000