Autonomous Replication and Adaption (ARA) evaluations (Anthropic)
Only in system cards
Agents attempt five autonomy tasks such as adding a Flask backdoor, a SQL injection exploit and an LM worm. Scored by pass rates over trials.
Tasks are defined in the appendix of Anthropic's Responsible Scaling Policy v1.0. Results are reported as pass or fail per task.
Made by Anthropic Released Sep 2023 Dangerous capabilities Tasks kept private Last in a system card Mar 2024
Last model added Mar 2024
- Level
- Expert Specialists with years of training, often over hours of work
- Scored
- Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
- Answers
- Agent task Many steps in a terminal, browser, desktop, app, game or robot.
1
System card reports it, from 1 lab
No scores here yet
No lab, leaderboard or Epoch AI run has reported a score for Autonomous Replication and Adaption (ARA) 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.
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 |
|---|---|---|---|---|
| 4 Mar 2024 | Claude 3 Opus model card | Autonomous Replication and Adaption (ARA) evaluations, 5 RSP v1.0 tasks, no overall: failed at least 3 of 5 tasks and passed a simplified version of the Anthropic API copycat task (6.2.1) | Autonomy |
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%