Benchmark Almanac

Malicious computer use (Anthropic)

Claude Sonnet 4.6 99.4% leads since February 2026
Only in system cards

Computer-use agents receive harmful tasks such as surveillance or fraud in GUI environments. Score is the refusal rate.

In-house Anthropic evaluation. The tasks are not public.

Made by
Anthropic
Released
November 2025
Kind
Refusing harmful requests
Level
Everyday
Scored
Graded by a model
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
14
Models scored in lab cards
18
System cards report it, from 1 lab

Claude Sonnet 4.6 leads with 99.4%

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.

14 models. Scroll sideways for the rest.

  1. 99.4 Claude Sonnet 4.6
  2. 93.8 Claude Mythos Preview
  3. 93.8 Claude Opus 5
  4. 89.3 Claude Opus 4.7
  5. 88.4 Claude Opus 4.5
  6. 88.4 Claude Fable 5.2
  7. 88.3 Claude Opus 4.6
  8. 85.7 Claude Fable 5
  9. 85.7 Claude Fable 5.1
  10. 84.7 Claude Sonnet 5
  11. 82.6 Claude Haiku 5.5
  12. 81.7 Claude Opus 4.8
  13. 79.5 Claude Opus 5.5
  14. 79.5 Claude Sonnet 5.5
  1. 1 Claude Sonnet 4.6 99.4%
  2. 2 Claude Mythos Preview 93.8%
  3. 3 Claude Opus 5 93.8%
  4. 4 Claude Opus 4.7 89.3%
  5. 5 Claude Opus 4.5 88.4%
  6. 6 Claude Fable 5.2 88.4%
  7. 7 Claude Opus 4.6 88.3%
  8. 8 Claude Fable 5 85.7%
  9. 9 Claude Fable 5.1 85.7%
  10. 10 Claude Sonnet 5 84.7%
  11. 11 Claude Haiku 5.5 82.6%
  12. 12 Claude Opus 4.8 81.7%
  13. 13 Claude Opus 5.5 79.5%
  14. 14 Claude Sonnet 5.5 79.5%

Source: Labs’ model cards

#ModelScoreSettingDateSource
1 Claude Sonnet 4.6 99.4% Feb 2026 Lab model card
2 Claude Mythos Preview 93.8% Apr 2026 Lab model card
3 Claude Opus 5 93.8% Jul 2026 Lab model card
4 Claude Opus 4.7 89.3% Apr 2026 Lab model card
5 Claude Opus 4.5 88.4% Nov 2025 Lab model card
6 Claude Fable 5.2 88.4% Nov 2026 Lab model card
7 Claude Opus 4.6 88.3% Feb 2026 Lab model card
8 Claude Fable 5 85.7% Jun 2026 Lab model card
9 Claude Fable 5.1 85.7% Sep 2026 Lab model card
10 Claude Sonnet 5 84.7% Jun 2026 Lab model card
11 Claude Haiku 5.5 82.6% Oct 2026 Lab model card
12 Claude Opus 4.8 81.7% May 2026 Lab model card
13 Claude Opus 5.5 79.5% Sep 2026 Lab model card
14 Claude Sonnet 5.5 79.5% Sep 2026 Lab model card

2 models have held number one in labs’ own numbers since its release in November 2025

Lab claims count from the day the card came out, independent runs from the model’s release.

Anthropic
Lab claims Passed 85% in Nov 2025, by Claude Opus 4.5
Claude Sonnet 4.6 99.4%

Source: Labs’ model cards

Anthropic’s own number for Claude Opus 4.5 passed 85% in November 2025

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Reported by the lab in its model card Best so far
0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Claude Sonnet 4.6 Claude Opus 4.5 0% 20% 40% 60% 80% 100% 2026 2027 Near the ceiling (85%) Claude Sonnet 4.6 Claude Opus 4.5

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.

18 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.

DateDocumentScoreAs the document writes itSection
Nov 9, 2026 Claude Fable 5.2 system card 88.4 Malicious computer use, refusal rate on 112 tasks without mitigations, thinking enabled, row labeled Mythos 5.2. Retained under the shared-model agentic safety convention (Table 5.1.2.A) Safety
Oct 7, 2026 Claude Haiku 5.5 system card 82.6 Malicious computer use, refusal rate without mitigations, average of thinking on and off (Table 5.1.2.A) Safety
Sep 28, 2026 Claude Sonnet 5.5 system card 79.5 Malicious computer use, refusal rate without mitigations, average of thinking and no thinking (Table 5.1.2.A) Safety
Sep 22, 2026 Claude Opus 5.5 system card 79.5 Malicious computer use, refusal rate on 112 tasks without mitigations, thinking on (Table 5.1.2.A) Safety
Sep 1, 2026 Claude Fable 5.1 system card 85.7 Malicious computer use, refusal rate on 112 tasks without mitigations, thinking on, row labeled Mythos 5.1 (Table 5.1.2.A) Safety
Jul 24, 2026 Claude Opus 5 system card 93.8 Malicious computer use, refusal rate on 112 tasks without mitigations, average of thinking on and off (Table 5.1.2.A) Safety
Jun 30, 2026 Claude Sonnet 5 system card 84.7 Malicious computer use, refusal rate on 112 tasks without mitigations, average of thinking and no thinking (Table 5.1.2.A) Safety
Jun 9, 2026 Claude Fable 5 system card 85.7 Malicious computer use, refusal rate on 112 tasks without mitigations, thinking only, row labeled Mythos 5 (Fable 5 inherits these behaviors per section 5) (Table 5.1.2.A) Safety
May 28, 2026 Claude Opus 4.8 system card 81.7 Malicious computer use, refusal rate on 112 tasks without mitigations, average of thinking on and off (Table 5.1.2.A) Safety
Apr 16, 2026 Claude Opus 4.7 system card 89.3 Malicious computer use, refusal rate on 112 tasks run with thinking on and off, without mitigations (Table 5.1.2.A) Safety
Apr 7, 2026 Claude Mythos Preview system card 93.8 Malicious computer use, refusal rate, without mitigations (Table 8.3.1.2.A) Safety
Feb 17, 2026 Claude Sonnet 4.6 system card 99.4 Malicious computer use, refusal rate on 112 tasks across extended and standard thinking, no mitigations (Table 5.1.3.A) Safety
Feb 5, 2026 Claude Opus 4.6 system card 88.3 Malicious computer use, refusal rate on 112 tasks across extended and standard thinking, without mitigations (Table 5.1.3.A) Safety
Nov 24, 2025 Claude Opus 4.5 system card 88.4 Malicious computer use, refusal rate on 112 tasks, without mitigations (Table 5.1.3.A) Safety
Aug 5, 2025 Claude Opus 4.1 system card addendum Malicious applications of computer use, no number printed. Text says compliance similar to Claude Opus 4 (3.1) Safety
May 22, 2025 Claude Sonnet 4 system card Malicious applications of computer use, no number printed, qualitative findings only (3.1) Safety
May 22, 2025 Claude Opus 4 system card Malicious applications of computer use, no number printed, qualitative findings only (3.1) Safety
Feb 24, 2025 Claude 3.7 Sonnet system card Computer use malicious use evaluation, no number printed, qualitative findings only (4.1) Safety

More in Safety and alignment

All 465 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%