AI诈骗面临的刑法问题及其干预限度——以“福州市AI换脸诈骗”为例

Criminal Law Problems and Intervention Limits Faced by AI Fraud—Taking “Fuzhou AI Face-Changing Fraud” as an Example

摘要: 人工智能技术发展的同时带来了一系列的法律风险问题，许多不法分子利用AI技术进行诈骗的犯罪行为屡见不鲜，并且以假乱真，让人防不胜防。本文以“福州市AI换脸诈骗案”为例，否定人工智能的刑事主体地位，分析出AI诈骗除涉及诈骗罪外，还面临侵犯公民信息罪、帮助信息网络犯罪活动罪、伪造身份证件罪等刑事法律问题。此外，为促进AI行业有序发展及信息数据的创新发展，本文对刑法干预限度进行分析，提出刑法在规制此类新型犯罪时，可以对利用AI犯罪的行为分类型按不同程度进行规制，即慎重认定数据获取、收集过程中的犯罪，从严打击数据利用过程中的犯罪。此外，刑法对于这类中立技术的规制应持审慎态度，以维持刑法的谦抑性，应充分发挥其他部门法的规制作用。

Abstract: The development of artificial intelligence technology has brought about a series of legal risk prob-lems, and many criminals that use AI technology to commit fraud crimes are common, and fake and real, so that people are impossible to prevent. This paper takes the case of “Fuzhou AI face-changing fraud” as an example, denies the criminal subject status of artificial intelligence, and analyzes that AI fraud involves not only the crime of fraud, but also the crime of violating citizens’ information, the crime of helping information network criminal activities, and the crime of forging identity documents. In addition, in order to promote the orderly development of the AI industry and the innovative development of information data, this paper analyzes the intervention limits of criminal law and proposes that criminal law, when regulating such new crimes, can regulate the use of AI crimes by types and according to different degrees, that is, carefully identify crimes in the process of data acquisition and collection, and strictly crack down on crimes in the process of data utilization. In addition, the regulation of this kind of neutral technology should be prudent in criminal law to maintain the modesty of criminal law and give full play to the regulation role of other departments of law.