REGREACT: Self-Correcting Multi-Agent Pipelines for Structured Regulatory Information Extraction

Extracting structured, machine-readable compliance criteria from regulatory documents remains an open challenge. Single-pass language models hallucinate structural elements, lose hierarchical relationships, and fail to resolve inter-document dependencies. We introduce $\text{textsc}RegReAct$, a self-correcting multi-agent framework that decomposes regulatory information extraction into seven specialized stages, each with an $\mathit{\text{Observe–Diagnose–Repair}}$ (ODR) loop that validates outputs against the source, correcting not only model hallucinations but also cross-reference errors in the regulations themselves. To ensure structural accuracy, $\text{textsc}RegReAct$ constructs a typed criterion graph; to ensure completeness, it resolves external dependencies by retrieving, summarizing, and embedding referenced legal content inline, producing self-contained outputs. Applying $\text{textsc}RegReAct$ to three EU Taxonomy Delegated Acts, we construct a dataset comprising 242 activities with over 4,800 hierarchical criteria, thresholds, and enriched source summaries. Evaluation against a GPT-4o single-pass baseline confirms that $\text{textsc}RegReAct$ outperforms it across all structural and semantic metrics. Code and data will be made publicly available: https://github.com/RECOR-Benchmark/RECOR