近亿条数据、300余项预警规则 揭秘浙江用AI抓贪官

Core Viewpoint - The Zhejiang Province's bidding supervision system utilizes AI technology to enhance anti-corruption efforts by identifying potential misconduct in the bidding process [1][3]. Group 1: System Features - The Zhejiang bidding supervision system includes six functional modules: comprehensive situation, industry supervision, monitoring analysis, early warning, verification and disposal, and data foundation [3][4]. - The system integrates data from 10 provincial departments, covering 12 categories of information related to investment projects and market entities, totaling nearly 100 million data entries [3][4]. - It has developed over 300 early warning rules to monitor industry irregularities and corruption, covering the entire bidding process from pre-bid to post-bid [4]. Group 2: Operational Impact - The system has shifted the approach to bidding supervision from passive to a combination of proactive and reactive measures [4]. - An example case involved the analysis of a project in Jiangshan City, where the AI system generated risk warnings about potential unfair practices, leading to a successful investigation by local authorities [5]. - The system has completed practical applications in 22 projects, resulting in the disqualification of 6 companies suspected of collusion, demonstrating its effectiveness in deterring misconduct [5].