Sora等生成式人工智能对诈骗犯罪司法认定的影响与应对

The Impact and Response of Generative Artificial Intelligence such as Sora on the Judicial Identification of Fraud Crimes

文生视频大模型Sora是目前生成式人工智能中最为先进的代表性应用。在为人类社会带来福祉及革命性变化的同时,Sora等生成式人工智能技术也有存在可能被滥用于进行“深度伪造”的潜在风险。在刑法层面,“深度伪造”等技术风险使得“眼见耳闻”不再为实,主要可能影响到诈骗类犯罪的司法认定与刑事归责。Sora等生成式人工智能现阶段仍属于弱人工智能,本质上属于人类使用的工具,不具有刑事责任主体资格。在未来的迭代版本中,我们不能排除Sora等生成式人工智能产生自主意识进而成为强人工智能的可能性。在Sora等生成式人工智能的介入下,司法机关将面临前所未有的证据认定难题。Sora等生成式人工智能技术风险的协同治理必须由国家、社会和企业等多方主体共同参与,在技术研发、推广应用等多环节建立健全监督评估机制,并且由行政法、刑法等多个部门法共同规制。

Sora is the most advanced representative application in generative artificial intelligence.While bringing benefits and revolutionary changes to human society,generative artificial intelligence such as Sora may also create a series of technical risks such as abuse and"deep forgery".At the criminal law level,technical risks such as"deep forgery"make"seeing and hearing"no longer true,which may mainly affect the judicial determination and criminal liability of fraud crimes.Generative artificial intelligence such as Sora is still weak artificial intelligence at this stage.It is essentially a tool used by humans and does not qualify as a subject of criminal responsibility.In future iterative versions,we cannot rule out the possibility that generative artificial intelligence such as Sora will develop autonomous consciousness and become strong artificial intelligence.With the intervention of generative artificial intelligence such as Sora,judicial organs will face unprecedented difficulties in identifying evidence.The collaborative management of risks in generative artificial intelligence technologies such as Sora must involve the participation of multiple entities such as the state,society and enterprises.A sound supervision and evaluation mechanism must be established in technology research and development,promotion and application,and must be governed by the laws of multiple departments such as administrative law and criminal law.common regulation.