Against the backdrop of rapid AI industry iteration, Shenzhen has taken the lead in deploying an intelligent case‑handling system covering the full spectrum of labor dispute cases. The system relies on robust computing power and integrates years of arbitration expertise to build a panoramic knowledge graph encompassing 111 dispute types. This innovation provides a solid foundation for the application of large language models (LLMs) and ensures comprehensive enablement across case filing, mediation, document review, court hearings, and ruling—drastically shortening the overall case cycle. The core algorithm adopts a “knowledge graph backbone plus chain-of-thought driving” approach, underpinned by stable operation of chips and other infrastructure, to sustain high‑efficiency data processing and logical reasoning. As a result, the filing stage benefits from multi‑dimensional intelligent verification, improving accuracy.
During mediation, the system deeply analyzes dispute focal points, providing real‑time access to relevant legal provisions and historical case references, assisting arbitrators in precise mediation—a demonstration of AI’s adaptability in complex judicial scenarios. In document review, automated information extraction by AI greatly boosts efficiency, saving arbitrators substantial time. At the hearing stage, the system intelligently generates investigation outlines, ensuring no key points are missed and showcasing the advantages of LLMs in dynamic decision‑making. Ruling generation accelerates the drafting of arbitration decisions while enabling rapid data retrieval. Moreover, the system breaks down data silos between arbitration and litigation, enhancing credibility and providing technical support for consistent rulings and case closure. This highlights AI’s potential in fostering judicial uniformity.
The system’s innovation has received high recognition from the Ministry of Human Resources and Social Security’s Mediation and Arbitration Department, reflecting the practical value of AI in addressing pain points of judicial services. Optimizations in chip‑level hardware have laid the foundation for computing power upgrades. Going forward, Shenzhen’s human resources and social security authorities plan to continuously refine system functions, expand deployment to city‑level arbitration bodies, build high‑quality datasets, and support nationwide mediation and arbitration efforts—further promoting the adoption of AI in public services in alignment with the smart transformation trend of the AI industry. This practice not only elevates arbitration service quality but also signals an accelerated convergence of AI and the judiciary; improvements in computing power, LLMs, and infrastructure will bring new opportunities for the industry.