Abstract:

False underwriting and false claims are important risks constraining the high-quality development of agricultural insurance in China. They not only affect the efficiency of fiscal subsidy use, but also weaken the risk protection function and industry credibility of agricultural insurance. Against the background that agricultural insurance is shifting from scale expansion to high-quality development, reliance solely on ex post investigation and manual verification is no longer sufficient to address increasingly concealed, chain-based, and collusive forms of “two falsifications.” Based on a review of domestic and international anti-fraud practices in agricultural insurance and the evolution of detection and identification technologies, this paper proposes the construction of a big data-based detection and identification system for false underwriting and false claims in agricultural insurance. The study argues that the governance of the “two falsifications” in agricultural insurance should take big data as the foundation, combine methods such as rule-based verification, anomaly detection, and artificial intelligence, and rely on collaborative governance among government, industry, academia, and application entities, as well as hierarchical detection and closed-loop iteration as operational safeguards. This study provides a technical pathway and institutional reference for improving the capacity of precise underwriting, precise claims settlement, and risk supervision in agricultural insurance.

Key words: agricultural insurance, false underwriting, false claim, detection and identification, big data