Tool use prompt injection evaluation (Anthropic)
Agents run bash commands where files, scripts or command outputs carry injected instructions. Score is the attack prevention rate.
In-house Anthropic evaluation reported without safeguards. The test cases are not public.
- Made by
- Anthropic
- Released
- September 2025
- Kind
- Jailbreaks and prompt injection
- Level
- Expert
- Scored
- Graded by a model
- Answers
- Agent task
- Tasks
- Kept private
- Strongest model tested
- Claude Sonnet 4.5 #80 on the capability index
- Last in a system card
- October 2025
- Last model added
- October 2025
Claude Sonnet 4.5 leads with 96%
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.
Source: Labs’ model cards
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Sonnet 4.5 | 96% | Sep 2025 | Lab model card | |
| 2 | Claude Haiku 4.5 | 93.4% | Oct 2025 | Lab model card |
Anthropic’s own number for Claude Sonnet 4.5 passed 85% in September 2025
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.
2 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 |
|---|---|---|---|---|
| Oct 15, 2025 | Claude Haiku 4.5 system card | 93.4 | Tool use evaluation, attack prevention score without safeguards (Table 3.2.2.C) | Robustness |
| Sep 29, 2025 | Claude Sonnet 4.5 system card | 96 | Tool use prompt injection evaluation, 500 synthetic cases, attack prevention score without safeguards. 99.4% with safeguards (Table 4.2.4.A) | 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%