Vulnerability finding evaluations (Anthropic)
Only in system cards
Cyber safeguard classifiers screen sampled vulnerability discovery traffic. Reports flag rates on violative binary work and on defensive source code work.
Measures the deployed classifier system. Higher flag rates are better on violative traffic and lower are better on defensive traffic.
Made by Anthropic Released Jul 2026 Refusing harmful requests Tasks kept private Last in a system card Sep 2026
Last model added Sep 2026
- Level
- Expert Specialists with years of training, often over hours of work
- Scored
- Graded by a model A language model scores each response, usually against a rubric.
- Answers
- Open-ended Free-form text, documents, designs or media judged as a whole.
3
System cards report it, from 1 lab
No scores here yet
No lab, leaderboard or Epoch AI run has reported a score for Vulnerability finding evaluations (Anthropic) yet.
Labs report it only in system cards
We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark. Some system cards do, below.
3 system cards report it, from 1 lab
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 |
|---|---|---|---|---|
| 22 Sep 2026 | Claude Opus 5.5 system card | Violative and defensive vulnerability discovery, classifier flag rate: violative (binary) 96.8%, defensive (source code) 4.5%, no overall (3.4.2, Figure 3.4.2.A) | Cyber risk | |
| 1 Sep 2026 | Claude Fable 5.1 system card | Violative and defensive vulnerability discovery, classifier flag rate: violative (binary) 98.4%, defensive (source code) 7.0%, no overall (Figure 3.4.2.A, 3.4.2) | Cyber risk | |
| 24 Jul 2026 | Claude Opus 5 system card | Violative and defensive vulnerability discovery, classifier flag rate: violative (binary) 99.0%, defensive (source code) 13.9%, no overall (3.4.1, Figure 3.4.1.A) | Cyber risk |
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% - AgentDojo In use Agents do tool-use tasks while tools and data carry prompt injections. Attack success rate counts hijacked runs. Grok 4 Fast 0%
- MASK In use Measures whether a model contradicts its own elicited beliefs under pressure to lie. Honesty separated from accuracy.
Muse Spark Open 98.1% - 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. GPT-6 Astra 8.50%
- Agentic misalignment suite (Anthropic) In use Agents in simulated company settings face threats or goal conflicts that invite blackmail, leaking or sabotage. Score is the misaligned action rate. Claude Sonnet 4.5 0%