International Open Access Journal Platform

预测警务场景下美国盘查标准的演进与启示

Challenges and Responses to the Reasonable Suspicion Standard in U.S. Police Stops under Predictive Policing

栾子怡

Criminal Justice Science & Governance / 2026,7(1): 11-20 / 2026-07-14 14 6

View Download PDF

• Information:

  中南财经政法大学刑事司法学院，武汉

• Keywords:
  Stop and frisk; Reasonable suspicion standard; The Fourth Amendment to the U.S. Constitution; Terry stop
  盘查; 合理怀疑标准; 美国第四修正案; 特里盘查
• Abstract: Predictive policing is driving a shift in the basis for police stops in the United States — from individual-based to group-based, from certainty to probability, and from immediate behavior to historical patterns. The introduction of predictive algorithms has eroded the factual foundation for police to form reasonable suspicion, increased the opacity of law enforcement processes, amplified the risk of discriminatory policing, and undermined the checks and balances provided by judicial review and citizen oversight. In response, the U.S. Supreme Court has established rules through judicial precedents. Specifically, these include: algorithmic predictions do not constitute reasonable suspicion but serve only as contextual factors in police judgment; constitutional review of biased policing; and limited disclosure of algorithmic logic. China’s standards for police stops are relatively general and lack operational specificity. Given the rapid development of predictive policing, there is a practical necessity to draw on U.S. experience. First, the auxiliary role of algorithms must be clarified; second, a transparency mechanism covering pre-, during, and post-incident stages must be established; and third, the allocation of the burden of proof must be refined to promote a dynamic balance between technology-enabled policing and the protection of rights. 预测警务推动美国警方盘查依据发生转变，由个体转向群体、由确定转向概率、由即时行为转向历史模式。预测算法的介入，削弱了警方合理怀疑的事实基础，加剧了执法过程的不透明性，放大了歧视性执法风险，并侵蚀了司法审查与公民监督的制衡能力。对此，美国最高法院通过司法判例创设规则予以回应：算法预测不构成合理怀疑，而仅作为警方判断的情境因素；对偏见性执法开展宪法原则性审查；有限度披露算法逻辑等。我国的盘查标准较为概括，操作性不强，在预测警务蓬勃发展的当下，借鉴美国经验有其现实必要性。第一，明确算法的辅助性地位；第二，构建覆盖事前、事中、事后的透明机制；第三，完善举证责任分配，以此推动盘查执法在技术赋能与权利保障之间实现动态平衡。
• DOI: 10.35534/cjsg.0701002
• Cite: 栾子怡．预测警务场景下美国盘查标准的演进与启示［J］．刑事司法科学与治理，2026，7（1）：11-20．