Every Great Tech Hub Needs Regulation
Hangzhou's courts are trying to keep generative AI in line—while encouraging its continued local development.
Trial scene in Hangzhou, Source
You might know Hangzhou, about an hour from Shanghai by high-speed rail, for its scenic West Lake or as the tech hub that nurtured DeepSeek.
But the city of over 12 million people is working on building a new reputation for itself: an innovator in AI regulation targeting areas such as copyright infringement and unfair competition.
In other words, Hangzhou wants to move fast and regulate.
In one case taken up by the Hangzhou Intellectual Property Court, “Xiaomoushu,” a reference to the popular lifestyle app “Little Red Book,” (the one American users flocked to during the TikTok ban in January) sued two companies for creating a tool that makes AI-generated posts in the style of the platform’s “notes.” The Plaintiff sued on the basis of both copyright infringement and unfair competition.
Judge Tang Xuebing, Source.
In August, the court announced it ruled in favor of “Xiamoushu,” finding that the generative AI product created unfair competition. Users, brands, and the platform itself would all be affected by a platform flooded with AI-generated marketing materials (some of which may have been fake or inaccurate).
Companies’ deployment of AI-generated material, Judge Tang Xuebing said, and “respect the rules of certain scenarios.” Firms should adhere to fair conduct and a “duty of care” in order to “prevent AI services from becoming tools of infringement.”
In other words: maybe think twice before releasing a generative AI product that’s going to ruin users’ experience of a platform.
Hangzhou’s model cases come in the wake of the city’s Implementation Plan for Accelerating the Development of Hangzhou as an Artificial Intelligence Hub, released in July. The plan is centered on boosting Hangzhou’s AI hardware and application ecosystem by directing subsidies to help attract talent and build out science parks, among other measures.
Judicial interpretations of the plan and Hangzhou’s broader AI development mandate indicate that judges’ belief that drawing clear red lines on AI deployment scenarios positively reinforces AI development. Hangzhou’s legal professionals will “both regulate industry development and stimulate innovation vitality,” said Chi Haijiang, Vice President of the Hangzhou Intermediate People’s Court.
Hangzhou IP Court’s case breakdown: patent cases dominate, but copyright-related cases are on the rise. Source.
Hangzhou authorities have good reason to worry about handling tech and AI-related suits. The city is an e-commerce hub and is home to Alibaba and Unitree Robotics, in addition to DeepSeek. Tech companies’ success is the city’s success: local officials “really care” about tech and innovation.
Local media connects the city’s tech reputation with its status as an early-mover on AI and copyright disputes. The careers of everyone involved rely on not interfering with (or appearing to interfere with) Hangzhou’s technological development. But that does not mean simply stepping aside and allowing for AI’s unruly deployment.
In order to preserve Hangzhou’s status as an innovation hub, courts are pursuing “judicial innovation” that both preserves rights holders’ protections and stimulates technological development.
As Chi from the Hangzhou Intermediate People’s Court put it: “The court will address technological uncertainty with judicial certainty.”
That means making full use of existent legal protections to establish frameworks for handling various types of AI-related cases that manage to protect them all: users, rights holders, and the enabling conditions for “innovation.” It’s important to note the courts choose which cases are publicized. It’s certainly possible that cases that do not offer neat solutions remain undisclosed.
Still, public discussion suggests support for legal interventions that might keep generative AI in check. Per an article in the Chinese newspaper 21st Century Business Herald, “Hangzhou is currently accelerating its development into an artificial intelligence innovation hub, and judicial services are also continuously innovating alongside technological changes.”
Talk of judicial innovation in handling IP and AI follows formal guidance directed at Hangzhou legal authorities over the summer. The “Opinions on Using High-Level Intellectual Property Judicial Services to Ensure Hangzhou’s Construction of an Artificial Intelligence Innovation Highland” (关于以高水平知识产权司法服务保障杭州建设人工智能创新高地的意见) directly link Hangzhou’s ability to meet the moment in court as essential to the city’s AI development ambitions.
The document’s release marks Zhejiang province’s first judicial protection opinion on AI. It calls for using existent laws such as the Anti-Unfair Competition Law in AI-related contexts, making sure perpetrators of AI-generated “misleading advertising,” for example, are punished. It also talks about reducing the burden of proof for rights holders, finding “market-based” dispute resolutions, and increasing judges’ AI expertise to form a “team of professional AI judicial experts.”
The Opinions are not a perfect blueprint to solve the innumerable potential cases stemming from new and controversial uses of generative AI. But Hangzhou’s experimentations could be worth watching for legal experts around the world who are reckoning with how to determine—and prove—wrongdoing in AI-related disputes.
An expert commentary on the Xiaomoushu case from Liu Wei of Tongji University explains that the defendants failed to exhibit reasonable care to “inform and remind” users that their product’s content was AI-generated—and failed to prevent that AI-generated content from being false and misleading. They not only harmed users, but also disrupted the genuine “ecosystem” of the “Little Red Book” app, thereby damaging market conditions.
“This case, for the first time, clarified the standards and boundaries for the commercial use of generative AI technology and provided a framework for adjudicating the duty of care of generative AI service providers,” he said.
It’s worth watching how closely those providers are paying attention to the court’s ruling. If they’re smart, they might be taking preliminary steps to avoid getting caught in its crosshairs.
What else I’m writing:
On Friday, I published a story in WIRED on accusations that Meta BitTorrented copyrighted porn to train its AI models.
“They have an interest in getting our content because it can give them a competitive advantage for the quality, fluidity, and humanity of the AI,” alleges Christian Waugh, an attorney for Strike 3.
Read the whole story here.