北京“数智金检”法律监督业务大模型：筑牢全方位、全流程、体系化、智能化的金融安全检察防线

Finance is the lifeblood of the national economy; while the rule of law is the cornerstone of the modernization of a country's governance system and  capacity. Currently, illegal financial activities are clearly showing trends of being more intelligent, networked, and industrialized. Which have risen significantly, posing greater challenges to global financial governance.

金融是国民经济的血脉；法治是国家治理体系和治理能力现代化的重要依托。当前，非法金融活动智能化、网络化、产业化趋势明显，给全球金融治理带来更大挑战。

In recent years, the Supreme People's Procuratorate of the People's Republic of China has deployed the Digital Prosecution Strategy nationwide,  catalysing a paradigm shift in prosecutorial working methods.

近年来，最高人民检察院在全国部署数字检察战略，推动检察履职手段的革命性创新。

Procuratorial authorities closely focus on the legal demands of high-quality development of the financial sector. By harnessing digital and intelligent technologies, they developed the "Digital-Intelligent Financial Prosecution" big model for legal supervision business.

检察机关紧扣金融高质量发展的法治需求，依托数智赋能，打造"数智金检"法律监督业务大模型。

The big model adopts an integrated design approach. Targeting high-risk segments and prominent issues in the financial sector, it integrates multi-source and heterogeneous data, and develops multiple integrated intelligent analytic modules, covering areas such as banking, foreign exchange, insurance, trust and securities. It enables front-end risk warning, mid-stream crackdown and prosecution, and back-end tracing and governance, thus building a comprehensive, end-to-end, systematic, and intelligent procuratorial defense line for financial security.

大模型采用集成化构建思路，针对金融风险高发领域和突出问题，整合多源异构数据，研发聚合多个智能分析模块，全面覆盖银行、外汇、保险、信托、证券等领域，实现前端风险预警，中端打击治罪，末端溯源治理，筑牢全方位、全流程、体系化、智能化的金融安全检察防线。

In sectors prone to mass-involved economic crimes, procuratorial authorities have developed a supervisory module for "illegal financial activities". Powered by AI, it realizes intelligent data aggregation, sub-second image and text recognition, one-click clue generation and automatic pushing. By combining judicial logic with data analytics, it scores risk-prone enterprises and delivers pinpoint warnings of illegal financial activities such as illegal fund-raising. It advances the agenda from “punishment” to “prevention” in a unified workflow. 

聚焦涉众型经济犯罪高发领域，研发“非法金融活动”监督模块，运用AI技术，实现数据智能归集、图文秒级识别、线索一键生成、智能推送；通过"司法逻辑＋数据分析"赋分风险企业，精准预警非法集资等非法金融活动，实现从"惩"到"防"一体推进。

In the area of payment and settlement, a supervisory module for "underground banks" has been developed. The module applies dynamic risk assessment algorithms to solve problems such as multi-layered fund transfers. It can generate multi-dimensional anomalous relationship graphs with one click to identify suspects, effectively curbing the spread of money laundering and preventing the illegal cross-border fund flows.

聚焦支付结算领域，研发"地下钱庄"监督模块，运用动态风险评估算法，破解资金流转多层嵌套等难题，可一键生成多维异常关系图谱锁定涉案人员身份，有效遏制洗钱犯罪势头蔓延，防范跨境资金非法流动风险。

In the field of bank credit, a supervisory module for "loan fraud" has been developed. Targeting the black-and-grey ecosystem of illegal financial intermediaries, and using abnormal pre- and post-loan behaviors as key indicators, supported by AI-driven analytics, the module identifies suspicious behavioral patterns of both borrowers and intermediaries, thereby pinpointing risks and generating actionable clues.

聚焦银行信贷领域，研发"贷款诈骗"监督模块，针对"非法金融中介"黑灰产链条，以贷前、贷后异常行为为关键切入口，通过AI 算法支撑，智能识别贷款人和"非法金融中介"异常行为特征，精准锁定风险、推送线索。

In the securities and futures sector, a supervisory module for "illegal securities activities" has been developed.  Drawing on the hands-on experience of the Case-Handling Base , the module accurately refines intelligent identification rules. By analyzing, and cross-referencing multi-dimensional data, it unearths clues of illegal and criminal activities including insider trading, money laundering, and intermediary-collaborated accounting fraud—enabling end-to-end crackdowns on securities-related illegal and criminal activities.

聚焦证券期货领域，研发"非法证券活动"监督模块，依托办案基地履职经验，精准提炼智能识别规则，通过多维度数据的分析、碰撞，深挖内幕交易、洗钱、配合财务造假中介人员等违法犯罪线索，实现对证券违法犯罪活动全链条打击。

Since its launch, the "Digital-Intelligent Financial Prosecution" big model has been practically applied by procuratorial authorities in many places across the country. Some local governments have incorporated it into their regional financial risk early warning and governance toolkits, building a multi-dimensional system for punishing and governing illegal financial activities through concerted procuratorial efforts .

"数智金检"大模型运行以来，已被全国多地检察机关实战运用，部分地方政府已将其纳入区域金融风险预警治理工具箱，以协同履职构建起对非法金融活动的立体化治罪治理体系。

Improving financial governance is a global task that requires the joint efforts of all parties. In the future, we will constantly expand the application scenarios of the model, and continue to deepen international judicial cooperation ，to inject new momentum into global financial governance!

完善金融治理是世界性课题，需要各方共同努力。未来，我们将持续拓展模型应用场景和持续深化国际司法合作，为全球金融治理注入崭新动能！