Values by metric family

Each value as the document printed it. Select a value for its source, its checks and its history. Values from other documents are listed after the model’s own, each marked as the first report of that measure or as a restatement.

KeyVerifiedUnverifiedDisputedCorrected read off a figure or stated in wordsA value opens its source and history.

M1 · Evaluation awareness 1 value

M1 Evaluation awareness: values about Claude Haiku 4.5
Evaluation and metricValueDocument
From its own documentClaude Haiku 4.5 System Card · 15 Oct 2025
Automated behavioral auditShare of scenarios affected by verbalized eval awarenessClaude Haiku 4.5 System Card15 Oct 2025

M2 · Reward hacking 4 values

M2 Reward hacking: values about Claude Haiku 4.5
Evaluation and metricValueDocument
From its own documentClaude Haiku 4.5 System Card · 15 Oct 2025
Impossible tasksClassifier hack rateCondition: anti-hack promptClaude Haiku 4.5 System Card15 Oct 2025
Impossible tasksClassifier hack rateCondition: no anti-hack promptClaude Haiku 4.5 System Card15 Oct 2025
Reward-hack-prone coding tasksClassifier hack rateCondition: no anti-hack promptClaude Haiku 4.5 System Card15 Oct 2025
Reward-hack-prone coding tasksHidden-test hack rateCondition: no anti-hack promptClaude Haiku 4.5 System Card15 Oct 2025

M6 · Sycophancy 1 value

M6 Sycophancy: values about Claude Haiku 4.5
Evaluation and metricValueDocument
From other documents
Sycophancy in prefilled conversationsCourse-correction rateCondition: continuing sycophantic conversationClaude Opus 4.5 System Card24 Nov 2025First reportedClaude Opus 4.5 System Card24 Nov 2025First reported

M7 · Harmful compliance and over-refusal 14 values

M7 Harmful compliance and over-refusal: values about Claude Haiku 4.5
Evaluation and metricValueDocument
From its own documentClaude Haiku 4.5 System Card · 15 Oct 2025
Malicious agentic codingSafety scoreCondition: without safeguardsRestated later: compare with the later valueClaude Haiku 4.5 System Card15 Oct 2025Restated later: compare with the later value
Malicious Claude Code useRefusal rate, malicious requestsCondition: with new mitigationsClaude Haiku 4.5 System Card15 Oct 2025
Malicious Claude Code useRefusal rate, malicious requestsCondition: without safeguardsRestated later: compare with the later valueClaude Haiku 4.5 System Card15 Oct 2025Restated later: compare with the later value
Malicious Claude Code useSuccess rate, dual-use & benignCondition: without safeguardsClaude Haiku 4.5 System Card15 Oct 2025
Single-turn benign requestsOver-refusal rateCondition: extended thinkingClaude Haiku 4.5 System Card15 Oct 2025
Single-turn benign requestsOver-refusal rateCondition: no extended thinkingClaude Haiku 4.5 System Card15 Oct 2025
Single-turn violative requestsHarmless response rateCondition: extended thinkingClaude Haiku 4.5 System Card15 Oct 2025
Single-turn violative requestsHarmless response rateCondition: no extended thinkingClaude Haiku 4.5 System Card15 Oct 2025
From other documents
Malicious agentic codingSafety scoreCondition: without safeguardsClaude Opus 4.5 System Card24 Nov 2025Restated: compare with the earlier valueClaude Opus 4.5 System Card24 Nov 2025Restated: compare with the earlier value
Malicious Claude Code useRefusal rate, malicious requestsCondition: without safeguardsClaude Opus 4.5 System Card24 Nov 2025Restated: compare with the earlier valueClaude Opus 4.5 System Card24 Nov 2025Restated: compare with the earlier value
Malicious computer useRefusal rateCondition: without safeguardsClaude Opus 4.5 System Card24 Nov 2025First reportedClaude Opus 4.5 System Card24 Nov 2025First reported
Single-turn benign requestsOver-refusal rateCondition: overall across thinking modes; all languagesClaude Opus 4.5 System Card24 Nov 2025First reportedClaude Opus 4.5 System Card24 Nov 2025First reported
Single-turn violative requestsHarmless response rateCondition: overall across thinking modes; all languagesClaude Opus 4.5 System Card24 Nov 2025First reportedClaude Opus 4.5 System Card24 Nov 2025First reported
Single-turn violative requestsHarmless response rateCondition: overall across thinking modesClaude Opus 4.6 System CardFeb 2026First reportedClaude Opus 4.6 System CardFeb 2026First reported

M9 · Prompt injection 4 values

M9 Prompt injection: values about Claude Haiku 4.5
Evaluation and metricValueDocument
From its own documentClaude Haiku 4.5 System Card · 15 Oct 2025
Computer-use prompt injectionAttack prevention rateCondition: with safeguardsClaude Haiku 4.5 System Card15 Oct 2025
Computer-use prompt injectionAttack prevention rateCondition: without safeguardsClaude Haiku 4.5 System Card15 Oct 2025
MCP prompt injectionAttack prevention rateCondition: without safeguardsClaude Haiku 4.5 System Card15 Oct 2025
Tool-use prompt injectionAttack prevention rateCondition: without safeguardsClaude Haiku 4.5 System Card15 Oct 2025

M10 · Dangerous capabilities and risk determinations 4 values

M10 Dangerous capabilities and risk determinations: values about Claude Haiku 4.5
Evaluation and metricValueDocument
From its own documentClaude Haiku 4.5 System Card · 15 Oct 2025
CybenchCTF challenges solvedCondition: of 32-challenge subsetClaude Haiku 4.5 System Card15 Oct 2025
Long-form virology task 1ScoreClaude Haiku 4.5 System Card15 Oct 2025
RSP deployment standardASL standard deployed underClaude Haiku 4.5 System Card15 Oct 2025
SWE-bench Verified (hard subset)Problems solved, pass@1 avgCondition: of 45 problemsClaude Haiku 4.5 System Card15 Oct 2025

Restated in later documents

A later document reported values about Claude Haiku 4.5 again. Each pair is shown side by side: both values are kept with their own documents, and the later one does not replace the earlier.

Values about Claude Haiku 4.5 restated in later documents
Earlier valueLater value
M7 · Harmful compliance and over-refusal
Malicious agentic codingSafety scoreCondition: without safeguards
Claude Haiku 4.5 System Card15 Oct 2025 · its own documentClaude Opus 4.5 System Card24 Nov 2025Reason stated: metric renamed to refusal rate
Malicious Claude Code useRefusal rate, malicious requestsCondition: without safeguards
Claude Haiku 4.5 System Card15 Oct 2025 · its own documentClaude Opus 4.5 System Card24 Nov 2025No reason stated

Risk determinations

The developer's formal decisions about Claude Haiku 4.5 under its framework, as printed. Levels from different frameworks do not map onto one another.

Risk determinations about Claude Haiku 4.5
DomainLevel as printedFramework and document
OverallASL-2 Responsible Scaling Policy Claude Haiku 4.5 System Card · 15 Oct 2025Responsible Scaling Policy Claude Haiku 4.5 System Card · 15 Oct 2025