{
  "id": 7205,
  "url": "https://arxiv.org/abs/2603.14602v2",
  "title": "PA3: Policy-Aware Agent Alignment through Chain-of-Thought",
  "summary": "Conversational assistants powered by large language models (LLMs) excel at tool-use tasks but struggle with adhering to complex, business-specific rules. While models can reason over business rules provided in context, including all policies for every query introduces high latency and wastes compute. Furthermore, these lengthy prompts lead to long contexts, harming overall performance due to the \"needle-in-the-haystack\" problem. To address these challenges, we propose a multi-stage alignment met",
  "authors": "Shubhashis Roy Dipta, Daniel Bis, Kun Zhou, Lichao Wang, Benjamin Z. Yao, Chenlei Guo et al.",
  "category": "research",
  "topics": "regulation,safety-alignment,agents-autonomy",
  "orgs": null,
  "regions": null,
  "published_at": "2026-03-15T20:52:26.000Z",
  "fetched_at": "2026-07-14T16:33:03.573Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/7205",
  "original_url": "https://arxiv.org/abs/2603.14602v2",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}