Beijing Adds 23 New Generative AI Services to Compliance Registry, Total Now at 128

Beijing’s cyberspace authorities have officially added 23 more generative AI services to the city’s compliance registry, bringing the total number of registered AI models and platforms to 128 as of April 10. This milestone reflects Beijing’s continued effort to enforce and encourage adherence to China’s generative AI governance framework, which went into effect in August 2023.

The update was published by the Beijing Municipal Bureau of Economy and Information Technology. Although the new list of approved services was not immediately disclosed, previously registered models include major offerings from companies like Baidu (ERNIE Bot), Alibaba (Tongyi Qianwen), iFlytek (SparkDesk), and Zhipu AI (GLM).

Under China’s AI service regulation, all public-facing generative AI products — including chatbots, content generators, and voice assistants — must undergo security assessments and be filed with local regulators. These requirements aim to ensure content safety, algorithm transparency, and protection of user data. Failure to comply may result in penalties or forced takedowns.

The rapidly expanding list highlights the fast-paced evolution of China’s generative AI landscape. Since registration first opened in August 2023, new entries have steadily increased month by month. The diversity of approved services — spanning sectors from education to enterprise automation — indicates that more startups and vertical SaaS platforms are launching LLM-powered features tailored to specific industry needs.

This regulatory environment also offers clues to China’s broader strategy: while promoting AI innovation, Beijing aims to maintain tight supervision over content generation and model behavior, especially as domestic alternatives to GPT-4 and Claude emerge.

With 128 registered services in Beijing alone — not including filings from other provinces like Shanghai, Guangdong, and Zhejiang — China’s regulatory model may become a blueprint for other countries grappling with responsible AI deployment.