AI firms penalized for fake product review scheme

Two generative artificial intelligence service providers were recently found to have engaged in unfair competition for using AI to mass-produce fake product endorsement posts on a major social media platform, according to a court judgment in East China's Zhejiang province.

The Hangzhou Intermediate People's Court ordered the service providers to pay 100,000 yuan ($14,723) in compensation to the social media platform, according to the court's WeChat account. The case, widely regarded as China's first involving unfair competition related to AI-generated content, has drawn broad attention on Chinese social media.

With the rapid development of AI-generative content technologies, a number of AI writing tools have appeared on Chinese social media, claiming to generate viral product recommendation posts with "just one click". At a very low cost, such tools can mass-produce fake consumer experiences and product endorsements.

In this case, heard by the Hangzhou court, the plaintiff was the operator of a well-known Chinese social media platform, referred to as Platform A, whose rules and community guidelines emphasize authentic experiences and genuine sharing.

The defendants, companies B and C, jointly operated an AI writing tool that could generate, with one click, product recommendation posts, travel guides and other content in the style of Platform A, while inducing users to publish such content on the platform.

The platform operator argued that the two companies had damaged its content ecosystem built around authentic product recommendations, and filed a lawsuit demanding that they cease copyright infringement and acts of unfair competition.

The AI writing tool's service page showed that its popular services included writing product recommendation posts, travel guides and post titles in the style of Platform A.

The service descriptions also claimed that the tool could directly generate posts "in line with the tone" of Platform A, or create "click-worthy titles" for users' posts on the platform.

By hitting a "one-click creation" button, users could directly generate posts of several hundred words suitable for publication on Platform A.

Zhang Zongbin, an assistant judge at the court, said the core issue in the case was whether providing targeted generative AI services for Platform A constituted unfair competition.

In the case, the service providers argued that their service was essentially a neutral technical tool.

However, Zhang said that while technology itself is neutral, that does not mean its application is neutral, noting that the key lies in where the technology is used, whom it targets and how it is applied.

"We believe it is necessary to prevent the abuse of AI technology as a cover to infringe upon the lawful rights and interests of others," he said.

During the trial, the court first clarified that the authentic content ecosystem built by Platform A through long-term operation constituted a legally protected competitive interest.

"Through long-term operation, Platform A invested substantial costs, manpower and resources, generated and accumulated a large volume of product recommendation content, and formed a content ecosystem centered on authentic experiences," Zhang said.

"This ecosystem is also a core resource through which the platform gains commercial benefits and competitive advantages. Therefore, we believe it is a competitive interest that can be protected under the Anti-Unfair Competition Law," he added.

Meanwhile, Zhang said the court mainly considered four factors before ordering companies B and C to compensate Platform A for economic losses and reasonable expenses.

These factors included whether the service in question constituted a generative AI service and whether it induced users to publish content not based on real experiences.

He also noted that the service was clearly driven by commercial interests, as the AI writing tool offered different levels of paid membership ranging from 40 yuan to 168 yuan.

From another perspective, Zhang said, the ruling clarified that generative AI service providers must observe responsibilities throughout the entire process, from content generation to compliance management.

Providers should, for example, clearly remind users of the limitations of AI-generated content, prohibit its use in publishing false information, and comply with the rules of specific platforms, he said.

In late April, the Office of the Central Cyberspace Affairs Commission, China's top internet regulator, launched a four-month campaign to address disorder in AI applications.

The campaign focuses on problems such as the use of AI to generate "digital garbage" and create or spread false information, according to the regulator.