Anthropic is launching its watermark verification API, letting approved organizations check whether text contains a digital watermark from Claude. The EU AI Act since August 2, 2025, requires new Claude models to embed invisible watermarks in their text output. "Regulators, law enforcement, media, fact-checkers, independent researchers, educational organizations, and EU civil society groups" can request access, as can enterprises that need to verify watermarking for their own compliance. Anthropic plans to expand access over time.
The system builds on Google's SynthID text method, tweaking word-selection randomness to create a statistically detectable pattern that, according to Anthropic, "may persist through some editing." That should make it far more reliable than traditional detectors like Pangram.
Anthropic says the watermark contains no user data and affects neither quality nor content. Critics disagree: if Claude picks synonyms based on a watermark key rather than meaning, text quality suffers. Trade publication "Artificial Lawyer" flags transparency risks too, since detectable AI fingerprints could become a problem where contracts ban AI use or during fee negotiations, though it predicts the watermarks are mostly harmless.
原始发布方：The Decoder：AI News（RSS）
原文时间：2026-09-02 04:40:36 +08:00
