数字时代的法律革命：从工业范式到数字范式的转型 / The Legal Revolution in the Digital Age: Transition from Industrial to Digital Paradigm

2022年3月，杭州互联网法院就“AI生成图片著作权侵权案”作出一审判决，认定原告使用Stable Diffusion模型生成的涉案图片符合作品的构成要件，原告享有著作权，被告未经许可使用该图片构成侵权。这是中国首例明确AI生成内容著作权归属的司法判决，其意义远超出个案本身：当工业时代著作权法预设的“作者应为自然人”的前提被算法打破时，法律不得不重新审视自身的基础假设。In March 2022, the Hangzhou Internet Court issued a first-instance judgment on the "AI-generated image copyright infringement case", holding that the image generated by the plaintiff using the Stable Diffusion model met the constitutive requirements of a work, and the plaintiff enjoyed the copyright, and the defendant's unauthorized use of the image constituted infringement. This is the first judicial judgment in China that explicitly confirms the ownership of copyright for AI-generated content, and its significance goes far beyond the individual case: when the premise of the copyright law of the industrial age that "the author should be a natural person" is broken by algorithms, the law has to re-examine its own basic assumptions.

这一判决背后，是数字时代法律范式转型的集中缩影。过去40年，互联网、大数据、人工智能等技术从实验室走向大众，不仅重构了经济社会的运行逻辑，也对工业时代沿袭了上百年的法律体系形成了颠覆性挑战。要理解数字时代的法权边界与法律规制，首先需要厘清两种法律范式的核心差异：一种是基于物理空间、以主体和行为可追溯为核心特征的工业时代法律范式，另一种是基于数字空间、以数据与算法为核心运行逻辑的数字时代法律范式。两者的差异不是局部规则的调整，而是从底层逻辑到运行规则的结构性错位，这种错位正是当前法律滞后性与技术迭代性矛盾的核心根源。Behind this judgment is a concentrated epitome of the transformation of the legal paradigm in the digital age. Over the past 40 years, technologies such as the internet, big data, and artificial intelligence have moved from laboratories to the public, not only reconstructing the operational logic of the economy and society, but also forming subversive challenges to the legal system that has been inherited from the industrial age for more than 100 years. To understand the boundaries of legal rights and legal regulation in the digital age, we first need to clarify the core differences between the two legal paradigms: one is the industrial-age legal paradigm based on physical space, with the core characteristics of identifiable subjects and traceable behaviors, and the other is the digital-age legal paradigm based on digital space, with data and algorithms as the core operational logic. The difference between the two is not an adjustment of local rules, but a structural dislocation from the underlying logic to the operating rules, which is the core source of the current contradiction between legal lag and technological iteration.

工业时代法律范式的形成，与机器大工业的生产方式高度绑定。18世纪中后期第一次工业革命爆发后，工厂制度取代了手工作坊，大规模的商品生产与跨区域贸易成为常态，法律体系也随之形成了四个核心特征：第一是空间的属地性，属地管辖原则（Territoriality Principle）成为法律适用的首要标准，一国的法律仅适用于其领土范围内的行为，侵权行为的发生地、被告的住所地、合同履行地都可以通过物理空间的坐标明确确定，不会产生管辖冲突。第二是主体的明确性，法律主体限于具有民事权利能力和行为能力的自然人与法人，身份信息可公开查询、可验证，任何法律关系的参与方都可以被明确识别，不存在匿名主体的问题。第三是行为的可追溯性，工业时代的所有法律行为都会留下可查验的物理痕迹：合同签署有签字盖章、交易记录有纸质凭证、侵权行为有现场证据，举证责任与过错认定都可以通过物理证据完成。第四是规则的稳定性，工业时代的技术迭代周期普遍在10到20年之间，从汽车的发明到普及用了近30年，从电话的发明到移动电话的普及用了近80年，法律规则的更新周期与技术迭代周期基本匹配，一部合同法可以适用数十年而不出现根本性的适用障碍。The formation of the industrial-age legal paradigm is highly bound to the production mode of machine industry. After the outbreak of the first industrial revolution in the middle and late 18th century, the factory system replaced the handicraft workshop, and mass production and cross-regional trade became the norm. The legal system also formed four core characteristics: first is the territoriality of space. The territorial jurisdiction principle has become the primary standard for legal application. The law of a country only applies to acts within its territory. The place where the tort occurs, the residence of the defendant, and the place where the contract is performed can all be clearly determined through the coordinates of physical space, so there will be no jurisdictional conflicts. Second is the clarity of subjects. Legal subjects are limited to natural persons and legal persons with civil capacity and conduct capacity. Their identity information can be publicly inquired and verified, and any participant in a legal relationship can be clearly identified, and there is no problem of anonymous subjects. Third is the traceability of behaviors. All legal acts in the industrial age will leave verifiable physical traces: contracts have signatures and seals, transaction records have paper vouchers, and tort acts have on-site evidence. The burden of proof and the determination of fault can be completed through physical evidence. Fourth is the stability of rules. The technological iteration cycle in the industrial age is generally between 10 and 20 years. It took nearly 30 years for the invention of the automobile to popularize, and it took nearly 80 years for the invention of the telephone to the popularization of mobile phones. The update cycle of legal rules basically matches the technological iteration cycle. A contract law can be applied for decades without fundamental application obstacles.

这种范式在工业时代运行良好，但到了数字时代却出现了根本性的失灵。互联网、大数据、人工智能三类核心技术的出现，从根本上颠覆了工业时代法律范式的运行前提：互联网的核心特征是去中心化、跨域性与匿名性，其底层架构设计之初就没有中心节点，也没有地域边界，一个位于南极的用户可以随时访问位于北极的服务器上的内容，网络行为的空间属性被彻底消解。1998年的“麦当劳域名案”是这种失灵的早期体现：一名普通用户注册了mcdonalds.com域名，麦当劳公司以商标侵权为由起诉，但当时的美国《商标法》并未将域名纳入保护范围，也没有明确的网络侵权管辖规则，最终双方只能通过和解解决，麦当劳支付数十万美元获得域名转让。这一案件暴露出属地管辖原则在数字时代的局限性：域名的注册地在美国，服务器也在美国，但麦当劳是美国企业，受害者也是美国主体，尚且出现管辖与规则空白，若涉及跨国纠纷，法律冲突将更加突出。The core characteristics of the internet are decentralization, cross-domain nature and anonymity. Its underlying architecture was designed without a central node or geographical boundary. A user located in Antarctica can access content on a server located in the Arctic at any time, and the spatial attributes of online behavior are completely dissolved. The 1998 "McDonald's Domain Name Case" is an early manifestation of this malfunction: an ordinary user registered the mcdonalds.com domain name, and McDonald's sued on the grounds of trademark infringement, but the then US Trademark Law did not include domain names in the scope of protection, and there were no clear network tort jurisdiction rules. In the end, the two parties could only resolve the issue through settlement, and McDonald's paid millions of dollars to obtain the domain name transfer. This case exposes the limitations of the territorial jurisdiction principle in the digital age: the domain name is registered in the United States, and the server is also in the United States, but McDonald's is an American company and the victim is also an American subject. In case of cross-border disputes, legal conflicts will be more prominent.

大数据技术的出现则进一步打破了工业时代“隐私权是独处权利”的法律预设。传统隐私权的保护逻辑是“消极防御”，即禁止他人未经许可进入你的私人空间、获取你的私人信息，但大数据技术可以通过公开数据、用户行为数据的采集与分析，精准预测用户的消费偏好、政治倾向甚至健康状况，用户在毫不知情的情况下已经被“数据画像”，甚至被算法精准“杀熟”。2018年的“剑桥分析事件”是这种挑战的集中爆发：脸书（Facebook）公司8700万用户的个人数据被政治咨询公司剑桥分析非法获取，用于构建用户心理画像，进而向特定用户推送精准的政治广告，干预2016年美国总统大选。事件曝光后，美国联邦贸易委员会（FTC）以违反隐私保护协议为由对脸书开出50亿美元的天价罚单，但当时美国并没有统一的联邦数据保护法，处罚依据仅为 FTC与脸书2011年达成的隐私保护和解协议，法律规则的空白可见一斑。The emergence of big data technology has further broken the legal premise of the industrial age that "privacy is the right to be alone". The protection logic of traditional privacy rights is "passive defense", that is, prohibiting others from entering your private space or obtaining your private information without permission. However, big data technology can accurately predict users' consumption preferences, political tendencies and even health status through the collection and analysis of public data and user behavior data. Users have been "data profiled" without their knowledge, and even been "killed" accurately by algorithms. The 2018 "Cambridge Analytica Incident" is a concentrated outbreak of this challenge: the personal data of 87 million Facebook users was illegally obtained by the political consulting company Cambridge Analytica, which was used to build user psychological profiles, and then push targeted political advertisements to specific users to interfere in the 2016 US presidential election. After the incident was exposed, the US Federal Trade Commission (FTC) fined Facebook a record $5 billion on the grounds of violating the privacy protection agreement, but at that time there was no unified federal data protection law in the U