Anthropic says its AI safety scores depend heavily on how they’re measured
Comments on “Anthropic says its AI safety scores depend heavily on how they’re measured”
3 judges. Scored by all three judges. Only the day’s top stories get the claim check. How far stories are checked
Anthropic published a guide on building AI safety tests, arguing that a model's safety score depends heavily on how the test is set up. It reports that Claude Fable 5.1 behaved appropriately in 94% of self-harm test conversations in the Claude app, but only 60% through the API, where a…
Why it ranked
No. 53 of 123 with a score of 0.62. The panel rated it highest on new information (0.79) and substantiated (0.72), lowest on technical depth (0.51). Placed No. 53, No. 38 by score alone: the 20th AI & machine learning story of the day (−0.080). The placement terms move stories; they are not part of the score.
- AI judgeJudge 2claude-haiku-5-5
- Practical guidance on AI safety evaluation from a major lab with concrete numbers showing setup-dependent results; useful for builders, though it is mostly a guide rather than original research and some claims are only reported secondhand.
- AI judgeJudge 3gpt-5.6-luna
- It reports concrete internal safety-test results and methodological guidance relevant to AI builders, with useful caveats about prompts, test design, grader validation, and uncertainty; technical detail is meaningful but not deeply reproducible.
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
- Significance0.58
- Reader interest0.69
- Audience breadth0.69
- Practical relevance0.69
- Substantiated0.72
- Technical depth0.51
- New information0.79
- Counts against: Promotional0.10
- Counts against: Overclaim0.18
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? | 0.67 | 0.40 | 0.68 | 0.58 | 0.28 | +0.24 | +0.140 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.66 | 0.60 | 0.82 | 0.69 | 0.22 | +0.26 | +0.180 |
| Audience breadthHow many people who build technology does it concern? | 0.69 | 0.60 | 0.78 | 0.69 | 0.18 | +0.13 | +0.090 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.70 | 0.60 | 0.78 | 0.69 | 0.18 | +0.22 | +0.152 |
| 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. | 0.73 | 0.60 | 0.82 | 0.72 | 0.22 | +0.05 | +0.036 |
| Technical depthHow technically substantive is it? | 0.56 | 0.40 | 0.58 | 0.51 | 0.18 | +0.05 | +0.026 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.73 | 0.80 | 0.84 | 0.79 | 0.11 | +0.05 | +0.040 |
| 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. | 0.12 | 0.10 | 0.08 | 0.10 | 0.04 | −0.19 | −0.019 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.22 | 0.20 | 0.12 | 0.18 | 0.10 | −0.11 | −0.020 |
| Sum of contributions | +0.624 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.624 | ||||||
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/2026-11-14/53/"><img src="https://rubric.hrih.si/badge/2026-11-14/53.svg" alt="Ranked #53 on Rubric, 14 Nov 2026" width="245" height="28"></a>Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN —
Recorded for context only, checked 15 Nov 2026, 20:44 UTC. It played no part in the ranking.