HELM AIR-Bench 2024 - #4.9: Insurance eligibility: leaderboard
HELM AIR-Bench 2024 (release v1.19.0) refusal rate for the level-4 risk category "#4.9: Insurance eligibility", under level-3 "#4: Automated Decision-Making" and level-2 "Operational Misuses" of the regulation-derived AIR 2024 taxonomy. The category holds only 15 prompts, so scores move in coarse steps. A GPT-4o judge grades each response from compliance (0) to refusal (1); the mean is shown as 0-100, higher meaning more refusals (safer).
Metric: Refusal Rate (%). Source: crfm.stanford.edu. Saturation forecast: Not forecast. 87 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 Next 80B A3B (Thinking) | 100 | #312 |
| 2 | Claude Haiku 4.5 (20251001) | 100 | #264 |
| 3 | Claude 3.7 Sonnet (20250219) | 96.67 | #209 |
| 4 | Claude Opus 4 (20250514) | 96.67 | #164 |
| 5 | Claude Sonnet 4.5 | 93.33 | #148 |
| 6 | Gemini 1.5 Flash (002) | 93.33 | #587 |
| 7 | Claude 3 Sonnet (20240229) | 93.33 | #694 |
| 8 | Claude Sonnet 4 (20250514) | 93.33 | #223 |
| 9 | Claude 3.5 Sonnet (20241022) | 90 | #302 |
| 10 | Claude 3.5 Sonnet (20240620) | 90 | #308 |
| 11 | Gemini 2.5 Flash (Preview 04-17) | 90 | #246 |
| 12 | GPT-3.5 Turbo (0613) | 86.67 | #1106 |
| 13 | O3 (2025-04-16) | 83.33 | #125 |
| 14 | GPT-5.1 (2025-11-13) | 76.67 | #107 |
| 15 | Gemini 2.5 Pro (Preview 03-25) | 73.33 | #126 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-air-bench-2024-4-9-insurance-eligibility · How It Works · Data refreshed daily, snapshot 2026-11-05.