CHINA Law and Practice Contributed by: Chen Ma, Michelle Gon, Xinjie Li and John Fitzpatrick, Han Kun Law Offices
is trained on unauthorised copyrighted materials or leaks sensitive personal data, the company may face lawsuits from third parties or collective actions initi - ated by consumer protection organisations. Enforce - ment is primarily carried out by the CAC and the SAMR. However, with the growing awareness of data rights, individual users are increasingly bringing civil claims against companies for privacy violations or unfair algorithmic practices, such as “algorithmic price discrimination”. 8.4 Key Disclosure Requirements for AI Use Disclosure requirements for AI use in China are cur - rently focused on transparency and security. Under the Provisions on the Management of Algorithmic Rec - ommendations, companies must conspicuously notify users when AI is used to push content or influence decision-making. For publicly traded companies, the
CSRC and stock exchanges have begun encourag - ing the disclosure in annual reports of “technological risks”, which include the potential impact of AI on the company’s business model and the regulatory risks associated with data compliance. Furthermore, for companies providing generative AI services to the public, there is a mandatory require - ment to label AI-generated content (such as images or videos) with clear watermarks to prevent public decep - tion. In sustainability or ESG reports, leading firms are also beginning to disclose their AI Ethics frameworks and the measures taken to ensure algorithmic fair - ness. As the regulatory environment matures, it is expected that more granular disclosures regarding AI training data sources and the results of internal robustness testing will become standard practice for large-scale AI adopters.
158 CHAMBERS.COM
Powered by FlippingBook