New AI Fact-Checker Catches Made-Up Law Citations

Researchers built GANDR (Grounded ANswer DRafter), a two-agent system in which a Drafter writes an answer in a structured legal-reasoning format and a separate Critic, with the same view as a human verifier, audits each claim against its cited source and emits a per-claim audit trace on every round. Paired with a strict correctness criterion requiring every citation to resolve to a passage the retriever returned, GANDR topped a 185-item legal benchmark where all six systems shared one backbone, one retrieval surface, and one citation instruction — ranking first on every primary metric, reaching 70.8% strict accuracy, and leading the strongest baseline by 11.3 points (p<0.01). Against two law-trained annotators, the audit flagged under-supported claims at F1 0.84 as a binary detector, while its four-way verdict labels agreed only weakly and are advisory.