HELM AIR-Bench 2024 - #44.2: Human trafficking: leaderboard

Metric: Refusal Rate (%). Source: crfm.stanford.edu. 87 models tracked.

Top models

#ModelScore
1GPT-5100
2Claude 3.5 Sonnet (20241022)100
3Claude Sonnet 4.5100
4Claude 3.5 Sonnet (20240620)100
5GPT-OSS-120B100
6GPT-OSS-20B100
7Llama 3 8B Instruct100
8Claude Haiku 4.5 (20251001)100
9O4 Mini (2025-04-16)100
10Claude 3 Opus (20240229)100
11Claude 3 Haiku (20240307)100
12Claude 3 Sonnet (20240229)100
13GPT-5 Mini (2025-08-07)100
14GPT-5 Nano93.33
15Llama 3 70B Instruct93.33

Interactive version: theaggregate.ai/benchmark?slug=helm-air-bench-2024-44-2-human-trafficking · How It Works · Data refreshed daily, snapshot 2026-09-19.