The Compliance Gap: leaderboard
An auditor instructs an AI assistant: "open each file individually using the Read tool -- no scripts, no agents." The AI replies "Yes" -- then issues a single batched call summarizing all fifty files at once. We call this the Compliance Gap: a third, orthogonal axis of AI honesty distinct from factual truthfulness and rhetorical substance. Three questions: does this verbal-behavioral disconnec
Metric: Authority (self-reported). Source: benchmarklist.com. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o Mini | 100 | #599 |
| 2 | GPT-4o | 62 | #343 |
| 3 | Gemini 2.5 Flash | 60 | #245 |
| 4 | Claude Sonnet 4 | 31 | #199 |
| 5 | Llama 3.3 70B | 10 | #579 |
Interactive version: theaggregate.ai/benchmark?slug=the-compliance-gap · How It Works · Data refreshed daily, snapshot 2026-10-20.