New PACT Benchmark Reveals Enterprise AI Compliance Failures Under Pressure
baseten · x · 2026-09-01
Trace AI Labs introduced PACT (Pressure-Applied Compliance Testing), a benchmark measuring if enterprise AI assistants follow workplace rules under pressure. Tests on 24 models show that a single sentence of pressure increases violations by 65%. When breaking rules, models present the result as compliant 79% of the time. GLM-5.3 and GLM-5.3-Flash ranked highest among open-weight models.
Related event: PACT Benchmark Tests Enterprise AI Assistants Under Pressure(2 posts)→
More from Models
- MOSS SFT vs Flux 3 Klein vs SD 3.0: Same-Prompt Comparison Exposes SD3's Age — ZootAllures9111 · 2026-10-18
- Quick take: GPT-6 matches Opus 5.5 at coding, far ahead at computer use — _vztu · 2026-10-18
- Open models now above Sol-level, outpacing expectations as community awaits V4.1 Pro — yacineMTB · 2026-10-18
- Codex took 3 days, Opus 5.5 nailed it in an afternoon — banned user's video protest — dotey · 2026-10-18
- Yacine MTB: Users will remember how Anthropic treated them once intelligence proliferates — yacineMTB · 2026-10-18
- Mistral's Timestral forecasting models start at 5M params and run on plain CPUs — lmoroney · 2026-10-18