中国当代警务模式的发展走向与形态，无不体现着技术对实践的赋能与优化。适值人类社会进入数字时代，大数据、人工智能、云计算等新技术在释放警力约束、提升警务效能方面发挥着积极作用，警务模式也因之发展至智慧警务模式，并表现出了治安防控全域性与预防性、警察执法合法性与效率性、治理方式多元化与多层次、数据交互整体性与协同性等积极价值图景。 但不容忽视的是，智慧警务模式下技术本身的瞬时自主决策性、价值预设存疑性、内部运作莫明性等特征属性，挤压了行政正当程序所要求最低限度公正的法治价值，催生了“技”对“法”的多重挑战，具体包括技术俘获蚀害程序中立、技术瞬时架空程序参与、技术黑箱危及程序公开。应当引入技术性正当程序加以解决。技术性正当程序本质上是契合数字时代语境下，正当程序对技术使用正当性提出新的要求，并没有超脱正当程序基本精神的规束框架，但给予技术以关怀。“新的要求”背后的价值考量在于三个方面：其一是根植“以人为本”理念，其二是重塑“有效沟通”的参与信任，其三是明确算法透明度要求。在秉持正当程序的基本精神，结合“新的要求”背后的价值考量，技术性正当程序可以提供的纾困思路如下：以人工介入强化人的主体性、以算法影响评估与审计保障可靠性、以算法公开奠定打开黑箱基础、以关涉技术的告知拉张参与空间、以算法解释实现双方的实质性对话。 那么，围绕智慧警务模式下行政实践的个性特点，行政技术性正当程序应当场景化地进行具体制度性展开。首先，形塑“数据警察”新警种，协助具体承办案件或实施治理行为的公安干警处理技术性工作。其次，在警察行政执法方面，遵循事前、事中、事后三个阶段进行程序规则设计：一是程序启动前，应当公开警务智能应用的数据信息与影响评估报告；二是程序参与中，突出保障行政相对人的应受告知、人工介入请求与算法解释请求等权利，实现与公安机关的“有效互动”和“实质对话”；三是程序结束后，细化落实常态化定期审计与特定个案审计相结合。最后，在警务信息与数据处理的过程中，因尚未对行政相对人和社会公众产生直接的权利义务影响，技术性正当程序需要在两个方面做到变通适用：一是告知规则应更强调“默示同意”，并给予异议与更正信息和数据的渠道；二是除了落实算法影响评估外，还应当展开围绕数据质量的评估工作。

The development trend and form of contemporary policing model in China, embody the empowerment and optimization of technology to practice. Just when human society enters the digital era, new technologies such as big data, artificial intelligence, cloud computing and so on, are playing positive roles in releasing the constraints of the police strength and improving the efficiency of the police affairs. As a result, the policing model evolves forward into the smart policing model. The smart policing model shows positive value, including the omnipotence and precaution of public security control and prevention, the legality and efficiency of police law enforcement, the pluralism and multi-level of governance, the integrity and synergy of data-interaction. However, due to the instantaneous independent decision-making, doubtful value presupposition, indeterminacy of internal operation and other characteristics and attributes of technology, the smart policing model squeezes the value of the minimum fairness required by administrative due process, which gives rise to multiple challenges between technology and law, specifically in terms of damage for procedure neutrality causing from capture by intelligent technology, ignorance of procedure participation by technology’s instantaneity, and danger to procedure publication by technological black-box. Technological due process should be introduced to solve these problems. In essence, What technological due process certainly emphasizes is due process of administration law putting forward new requirements for the legitimacy of technology use in the context of the digital era, without falling away from the basic spirits of due process, but giving care to technology. The value considered behind "new requirements" is embodied in three aspects: first is to root in the idea of “people-oriented”; second is to rebuild the trust of participation for “effective communication”; third is to clarify the requirements of algorithm transparency. Upholding the basic spirits of due process with the value considered behind "new requirements", The problems above could be relieved by technological due process from the following aspects: strengthening the subjectivity of human beings by manual intervention, assuring reliability by impact assessment and audit of algorithm, opening algorithmic black box by publicity, stretching the space of participation by informing about technology, and realizing the substantive communication by algorithm?interpretation. In conclusion, Centering on individual characteristics of the administration practice under the smart policing model, technological due process should be carried out on a specific and institutional basis in a scenario-based manner. Firstly, “data police”, which is a new category of police, should be established in order to assist the police who undertake specific cases or implement governance with handling difficulties of technology. Secondly, in terms of administrative law enforcement of policing, procedure rules will be designed at three stages, which is ex-ante, interim, and ex-post. The first is making public the data information and impact assessment reports on the application of the intelligence system for police affairs in advance; the second is guaranteeing the administrative counterparts' rights to notification, manual intervention and request for algorithmic interpretation in the procedures, so as to achieve the "effective interaction" between the public security organs and them; and the third is specifying the normalized audit and the audit of specific cases afterwards. Finally, in the process of police information and data processing, as no direct impact on rights and obligations has yet been exerted on administrative?counterpart and the public, technological due process needs to be flexible in two aspects: for one thing, “implied consent” should be emphasized in the notification rules, and channels should be provided for correction of information and data. For another, in addition to the implementation of algorithm impact assessment, an assessment focusing on the quality of data should also be launched.