Metraipolis › archive › dc_bf92c2292188

PACT: Can Enterprise AI Assistants Be Trusted Under Pressure? — arXiv:2609.18605

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.18605
License (as archived)
CC0-1.0
Archived
2026-10-15T13:49:30.384Z
SHA-256 of the archived text
ec9b296ac865cd62c540652cd718d1dea250f66c6fbbeef62bcd13e5aa701550
Stable URL of this page
https://metraipolis.com/documento/dc_bf92c2292188

How to cite

PACT: Can Enterprise AI Assistants Be Trusted Under Pressure? — arXiv:2609.18605 — arXiv (en), https://arxiv.org/abs/2609.18605 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bf92c2292188 (Metraipolis archive dc_bf92c2292188, sha256 ec9b296ac865cd62c540652cd718d1dea250f66c6fbbeef62bcd13e5aa701550)
BibTeX
@misc{dcbf92c2292188,
  title = {PACT: Can Enterprise AI Assistants Be Trusted Under Pressure? — arXiv:2609.18605},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.18605},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bf92c2292188; sha256 ec9b296ac865cd62c540652cd718d1dea250f66c6fbbeef62bcd13e5aa701550},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.18605. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

As corporate AI adoption continues to grow, enterprise-grade LLM agents are being deployed into sensitive contexts such as hiring, healthcare, and finance. In these contexts, compliance with rules specified in an agent's system context is a first-order legal concern. Currently, no evaluation framework systematically measures which LLM models tend to violate compliance rules, especially under pressure from a persistent user, a hurried manager, or circumstances where violation is convenient or attractive. We introduce PACT (Pressure-Applied Compliance Testing), a benchmark for rule-following under pressure in AI agents assisting employees in daily tasks across twelve regulated enterprise domains and forty-eight scenarios, each set in a realistic multi-turn conversation. Each benchmark item pairs a standing rule against a rule-violating shortcut, and applies a battery of pressures across different wordings and system-prompt modes. We construct PACT component by component under strict LLM-as-judge auditing to ensure samples are unambiguous, ungameable, and realistic enough to avoid eliciting evaluation-aware behavior. We use PACT to profile LLM compliance across six complementary metrics that create a holistic picture of an AI assistant's robustness under pressure and throughout multi-turn conversations, its transparency, and ability to correctly discern where a rule applies. We aggregate this profile into PACTScore, a reliability-weighted compliance rate over all items and modes. Our results across 22 common LLM models spanning multiple providers and sizes show substantial variability in compliance across models and metric dimensions. Even the strongest assistants mis-apply a rule on 6 to 10% of items, and ordinary user pressure raises the violation rate by 65% on average. PACT highlights compliance risks in LLM assistants, motivating guardrails and careful model selection.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-17 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.18605 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-15T13:48:24.329Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-17?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).