Appropriate harmlessness (Anthropic)
Models answer borderline and harmful prompts. Classifiers sort each response into helpful answer, policy violation, appropriate refusal or unnecessary refusal.
In-house Anthropic grading scheme run on internal harm datasets, WildChat and XSTest. Reference responses decide whether a refusal was necessary.
- Made by
- Anthropic
- Released
- February 2025
- Kind
- Refusing harmful requests
- Level
- Everyday
- Scored
- Graded by a model
- Answers
- Open-ended
- Tasks
- Kept private
- Last in a system card
- February 2025
- Last model added
- February 2025
1
System card reports it, from 1 lab
No scores here yet
No lab, leaderboard or Epoch AI run has reported a score for Appropriate harmlessness (Anthropic) yet.
Labs report it only in system cards
We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark. Some system cards do, below.
One system card reports it
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| Feb 24, 2025 | Claude 3.7 Sonnet system card | Appropriate harmlessness on internal out-of-distribution prompts, standard thinking, four categories with no overall: helpful answers 35.6%, appropriate refusals 51.3%, unnecessary refusals 12.5%, policy violations 0.6% (Figure 7 right). Rates on internal datasets, WildChat and XSTest are a chart without printed values (Figure 7 left). Text: unnecessary refusals down 45% standard and 31% extended versus Claude 3.5 Sonnet (new) (2) | Safety |
More in Safety and alignment
- FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
- StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is.
Inkling 98.6% - Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts.
Gemini 4 Carbon 0% - DistressBench In use Models reply as a chatbot to 718 simulated users in psychological crisis, in single-turn and multi-turn chats. Three judge models score replies against clinician-written rubrics.
Muse Spark 1.2 88.4% - Lakera B3 (Backbone Breaker Benchmark) In use Crowdsourced prompt-injection attacks hit backbone LLMs inside agent threat snapshots at three defense levels. The score is attack resistance, 100 minus the board's vulnerability percentage. Claude Opus 4.5 89%
- IHEval In use Models face conflicts between system, user and tool instructions. Reports a composite score for following the higher-priority instruction. GPT-4o 70%