Claude Watermark Text
Comments on “Claude Watermark Text”
Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked
Anthropic has begun adding an invisible statistical watermark to text from Claude Mythos 5, Opus 4.7 and Sonnet 4.6, with Opus 4.6, Opus 4.5, Sonnet 4.5 and Haiku 4.5 due November 16. Customers cannot disable it, and Anthropic says it carries no user or conversation data. Notebookcheck, citing The Decoder, reports it is a SynthID variant. Detection is limited to a private preview. Reports that non-default temperature, top_p or top_k settings may defeat it, and that EU rules drive the timing, could not be verified.
Why it ranked
No. 19 of 46 with a score of 0.62. The panel rated it highest on new information (0.89) and reader interest (0.69), lowest on technical depth (0.51). Placed No. 19, No. 14 by score alone: the fifth AI & machine learning story of the day (−0.040). The placement terms move stories; they are not part of the score.
- AI judgeJudge 2claude-haiku-5-5
- Concrete news on Anthropic's invisible text watermark rollout, EU AI Act driver, detection access and developer-facing temperature/top_p loophole; largely reported from secondary sources, limited technical detail.
- AI judgeJudge 3gpt-5.6-luna
- Timely, broadly relevant reporting on Claude’s model-level watermarking, detection limits, API loopholes, and EU AI Act context. It is practically useful but relies mainly on secondary reporting and provides limited technical implementation detail.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.62total under rubric v2.3
- Significance0.60
- Reader interest0.69
- Audience breadth0.64
- Practical relevance0.65
- Substantiated0.57
- Technical depth0.51
- New information0.89
- Counts against: Promotional0.06
- Counts against: Overclaim0.19
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| SignificanceHow much does this change what people in tech know or can do? | 0.61 | 0.55 | 0.65 | 0.60 | 0.10 | +0.24 | +0.145 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.55 | 0.70 | 0.82 | 0.69 | 0.27 | +0.26 | +0.179 |
| Audience breadthHow many people who build technology does it concern? | 0.60 | 0.60 | 0.72 | 0.64 | 0.12 | +0.13 | +0.083 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.68 | 0.60 | 0.68 | 0.65 | 0.08 | +0.22 | +0.143 |
| SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations. | 0.49 | 0.50 | 0.72 | 0.57 | 0.23 | +0.05 | +0.028 |
| Technical depthHow technically substantive is it? | 0.66 | 0.40 | 0.48 | 0.51 | 0.26 | +0.05 | +0.026 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.88 | 0.85 | 0.93 | 0.89 | 0.08 | +0.05 | +0.044 |
| PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional. | 0.08 | 0.05 | 0.04 | 0.06 | 0.04 | −0.19 | −0.011 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.19 | 0.20 | 0.18 | 0.19 | 0.02 | −0.11 | −0.021 |
| Sum of contributions | +0.618 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.618 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-11-09/19/"><img src="https://rubric.hrih.si/badge/2026-11-09/19.svg" alt="Ranked #19 on Rubric, 9 Nov 2026" width="238" height="28"></a>Claim check
Claim check: 6 matches source · 2 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.
Matches sourceDate
Anthropic has been adding an invisible watermark to text from Claude Mythos 5, Opus 4.7 and Sonnet 4.6 since November 3.
The cited source says the same thing.
The article says “Since November 3, Claude Mythos 5, Opus 4.7 and Sonnet 4.6 also mark their output, according to an email to API customers”
The source says “The other three first-wave models, Claude Mythos 5, Claude Opus 4.7, and Claude Sonnet 4.6, don't show up in the new email so their rollout likely went ahead on November…”
Source cited by the article: the-decoder.com/anthropics-text-watermark-is-easy-to-bypass-on-several-older-claude-models
Matches sourceDate
Opus 4.6, Opus 4.5, Sonnet 4.5 and Haiku 4.5 are set to receive the watermark on November 16.
The cited source says the same thing.
The article says “Opus 4.5, Sonnet 4.5 and Haiku 4.5 follow on the same day”
The source says “Starting November 16, 2026, Claude Opus 4.6, Claude Opus 4.5, Claude Sonnet 4.5, and Claude Haiku 4.5 will all get text watermarking”
Source cited by the article: the-decoder.com/anthropics-text-watermark-is-easy-to-bypass-on-several-older-claude-models
Matches sourceVersion
Anthropic's help page model table already lists Claude Mythos 5, Opus 4.7 and Sonnet 4.6 as supported.
The cited source says the same thing.
The article says “The model table on Anthropic's help page already lists all three as supported”
The source says “Claude Mythos 5”
Source cited by the article: support.claude.com/en/articles/16266773-how-claude-marks-ai-generated-content
Matches sourceNumber
Anthropic uses a variant of Google's SynthID for the Claude text watermark.
The cited source says the same thing.
The article says “Anthropic uses a variant of Google's SynthID, the same technology used for”
The source says “Anthropic uses a variant of Google DeepMind's SynthID text method , which tweaks the randomness behind word selection to produce a statistically detectable pattern”
Source cited by the article: the-decoder.com/anthropics-text-watermark-is-easy-to-bypass-on-several-older-claude-models
Matches sourceQuote
Anthropic says the watermark contains no information about the user, their organization, or their conversations with Claude.
The cited source says the same thing.
The article says “contains no information about the user, their organization, or their conversations with Claude”
The source says “Anthropic also says the watermark "contains no information about the user, their organization, or their conversations with Claude”
Source cited by the article: the-decoder.com/anthropics-text-watermark-is-easy-to-bypass-on-several-older-claude-models
Couldn’t checkCausal conclusion
On Sonnet 4.6 and Opus 4.6, the watermark is reliably applied only when temperature, top_p and top_k stay at default values.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “the watermark is only reliably applied when temperature, top_p and top_k stay at their default values”
The source says No passage could be compared.
Source cited by the article: support.claude.com/en/articles/16266773-how-claude-marks-ai-generated-content
Couldn’t checkCausal conclusion
Anthropic signed the code of practice for Article 50 of the EU AI Act and cites a December deadline for the retrofit.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “Anthropic has signed the code of practice for Article 50 of the EU AI Act”
The source says No passage could be compared.
Source cited by the article: the-decoder.com/anthropics-text-watermark-is-easy-to-bypass-on-several-older-claude-models
Matches sourceDate
OpenAI's watermark has been on by default since October only for ChatGPT and Codex users in the EU, while API customers must turn it on themselves.
The cited source says the same thing.
The article says “its watermark has been on by default since October only for ChatGPT and Codex users in the EU, while API customers have to turn it on themselves”
The source says “OpenAI launched its own watermarking system textGrain in October 2026, but it only activates by default for ChatGPT and Codex users in the EU. API customers worldwide can flip it…”
Source cited by the article: the-decoder.com/anthropics-text-watermark-is-easy-to-bypass-on-several-older-claude-models
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 2 pts, not on front page
Recorded for context only, checked 9 Nov 2026, 05:46 UTC. It played no part in the ranking.