Structured Decomposition for Reliable LLM-Generated Access Control Policies
An LLM system translates natural-language access control policies into executable Rego code with 50.3% correctness rate.
Researchers present an LLM-based system that translates natural-language access control policies into executable Rego code for Open Policy Agent. The system uses structured decomposition and schema-aware validation to achieve a 50.3% end-to-end correctness rate, a 3.3x improvement over a direct single-prompt baseline. It demonstrates significant improvements in generating correct deny semantics and ownership conditions.
- The system translates natural-language access control policies into executable Rego code for Open Policy Agent.
- It achieves 50.3% end-to-end policy correctness, a 3.3x improvement over baseline.
- Structured decomposition and schema-aware validation significantly improve LLM-generated policy reliability.
Full article202 words · extracted from arxiv.org · click to collapse
This paper presents an LLM-based system that translates natural-language access control policies (NLACPs) into executable Rego code for Open Policy Agent (OPA). It provides a modular, end-to-end pipeline for policy detection, component extraction, schema validation, linting, compilation, and automated test generation and execution. The system is designed to bridge the gap between human-readable access requirements and machine-enforceable policy-as-code (PaC), with a focus on deployment reliability and security correctness. We evaluate the system on 372 ACRE-complete access control statements with non-null subject, action, and resource annotations against a direct single-prompt LLM baseline to isolate the contribution of structured decomposition and schema-aware validation. The system achieves a 50.3% end-to-end policy correctness rate, compared with 15.3% for the baseline, representing a 3.3x improvement. A policy is counted as correct only if it satisfies compilation, linting, and both positive and negative tests, making this a strict measure of deployable correctness. On security-critical patterns, the system generates correct deny semantics for 87.5% of deny policies (baseline: 37.5%), ownership conditions for 100% of ownership-qualified policies (baseline: 40%), and status-qualified conditions for 100% of status-qualified policies (baseline: 55.6%). These results indicate that structured decomposition and schema-aware validation play a critical role in improving the reliability of LLM-generated authorization policies.
Text extracted automatically; images, tables and formatting may be missing. Original: https://arxiv.org/abs/2609.24036