Code-as-Auditor: Executable Compliance Reasoning via Regulation-to-Code

arXiv:2609.19199v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly adopted for compliance and legal reasoning tasks, yet their outputs often lack explicit grounding in legal logic and evidence. We present…

Read the full story at arXiv cs.AI ↗