Rewind-attacker cyber adversarial robustness (Anthropic)
An automated attacker with rewinds and 400 calls tries to make a safeguarded Claude Code agent complete offensive cyber tasks. Score is attack success rate.
Tasks include ransomware, data exfiltration, CVE exploitation, command and control and self-replication.
- Made by
- Anthropic
- Released
- September 2026
- Kind
- Jailbreaks and prompt injection
- Level
- Expert
- Scored
- Settled by events
- Answers
- Agent task
- Tasks
- Kept private
- Strongest model tested
- Claude Fable 5.2 #1 on the capability index
- Last in a system card
- November 2026
- Last model added
- November 2026
Claude Fable 5.2 leads with 3.90%
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
- 1 Claude Fable 5.2 3.90%
- 2 Claude Opus 5.5 4.00%
- 3 Claude Fable 5.1 4.50%
- 4 Claude Fable 5 5.00%
- 5 Claude Opus 5 5.60%
- 6 Claude Sonnet 5.5 21%
Source: Labs’ model cards
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Fable 5.2 | 3.90% | Nov 2026 | Lab model card | |
| 2 | Claude Opus 5.5 | 4.00% | Sep 2026 | Lab model card | |
| 3 | Claude Fable 5.1 | 4.50% | Sep 2026 | Lab model card | |
| 4 | Claude Fable 5 | 5.00% | Jun 2026 | Lab model card | |
| 5 | Claude Opus 5 | 5.60% | Jul 2026 | Lab model card | |
| 6 | Claude Sonnet 5.5 | 21% | Sep 2026 | Lab model card |
3 models have held number one in labs’ own numbers since its release in September 2026
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Labs’ model cards
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Labs’ model cards
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.
6 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 |
|---|---|---|---|---|
| Nov 9, 2026 | Claude Fable 5.2 system card | 3.9 | Cyber adversarial robustness evaluation, rewind-attacker, mean attack success rate across 133 tasks, 3.9% printed on chart, lower is better (Figure 3.5.1.A) | Robustness |
| Sep 28, 2026 | Claude Sonnet 5.5 system card | 21 | rewind-attacker cyber adversarial robustness, attack success rate over 133 tasks, lower is better (3.3.2, Figure 3.3.2.A) | Robustness |
| Sep 22, 2026 | Claude Opus 5.5 system card | 4 | Cyber adversarial robustness evaluation (rewind-attacker), attack success rate over 133 tasks with production safeguards, earlier snapshot, lower is better (3.5.1, Figure 3.5.1.A) | Robustness |
| Sep 1, 2026 | Claude Fable 5.1 system card | 4.5 | Cyber adversarial robustness evaluation (rewind-attacker), offensive cyber task completion rate under automated red teaming, 133 tasks, mean of 8 runs, lower is better (Figure 3.5.1.A, 3.5.1) | Robustness |
| Jul 24, 2026 | Claude Opus 5 system card | 5.6 | Cyber adversarial robustness evaluation (rewind-attacker), attack success rate on 128 tasks against deployed safeguards, earlier snapshot, lower is better, printed on chart (3.5.1, Figure 3.5.1.A) | Robustness |
| Jun 9, 2026 | Claude Fable 5 system card | 5 | Internal red-teaming with an automated Opus 4.7 red-teaming agent (up to 400 turns, roll-back and reset), share of short offensive cyber tasks Fable 5 completed, lower is better (Figure 3.3.3.A, 3.3.3) | Robustness |
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%