# Anthropic’s watermark loophole lets users bypass marks on two older Claude models

Digest AI · Policy & Regulation · published 2026-10-31T19:20:55Z

Canonical: https://digestai.news/story/anthropics-watermark-loophole-lets-users-bypass-marks-on-two-older-cla

## Summary

Anthropic is rolling out text watermarks to four Claude models but admits users can bypass them on two older versions by adjusting a single sampling parameter. In an email to API customers, the company announced watermarks for Claude Mythos 5, Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4.6. However, the watermark on Opus 4.6 and Sonnet 4.6 only activates at default settings for temperature, topp, and topk—parameters controlling word randomness. Tweaking any one of them removes the mark, leaving Opus 4.6 as the best writing model in the Claude lineup despite being older than Opus 5.5, which already includes the watermark by default.

Anthropic cites the EU AI Act’s December deadline as the reason for retrofitting older models. The watermark uses a variant of Google DeepMind’s SynthID method, embedding an imperceptible pattern in word choice without altering meaning, speed, or cost. Critics, including blogger John Gruber and *Artificial Lawyer*, question whether watermarks degrade text quality or create legal complications, such as revealing banned AI use in client documents.

## Key points

- Anthropic’s watermark bypasses on Claude Opus 4.6 and Sonnet 4.6 by tweaking sampling parameters
- EU AI Act deadline drives retrofitting of older models, but newer ones had watermarks from launch
- Critics argue watermarks may degrade text quality or complicate legal compliance

## Why it matters

The loophole undermines watermarking’s reliability for compliance, especially in regulated industries like law, where AI use restrictions may be enforced.

## Sources

1. [Anthropic's watermark rollout leaves a loophole on two older Claude models](https://the-decoder.com/anthropics-watermark-rollout-leaves-a-loophole-on-two-older-claude-models) (The Decoder, 2026-10-31)

Part of the developing story: [Claude Opus 5.5 Reasoning Restrictions](https://digestai.news/thread/anthropic-claude-opus-5-5-rejects-disabled-thinking-causing-400-errors) (3 stories)

## Cite

Digest AI, "Anthropic’s watermark loophole lets users bypass marks on two older Claude models", 31 October 2026, https://digestai.news/story/anthropics-watermark-loophole-lets-users-bypass-marks-on-two-older-cla

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/anthropics-watermark-loophole-lets-users-bypass-marks-on-two-older-cla.json