NDA non-compliance clause detection on Internal NDA 40-policy NDA-3
95AccuracyPolicyGuard
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| PolicyGuardModel=GPT-4.1, Protocol=zero-shot2026.06 | 95 | 85.7 | |
| PolicyGuardModel=Sonnet-4.5, Protocol=zero-shot2026.06 | 86.8 | 61.5 | |
| Claude CoworkModel=Opus-4.6, Guidelines=raw policy guidelines, Protocol=zero-shot2026.06 | 85 | 50 | |
| Claude CoworkModel=Sonnet-4.6, Guidelines=raw policy guidelines, Protocol=zero-shot2026.06 | 72.5 | 35.3 |