China's Cyberspace Administration registered 72 new generative AI services from March to April, bringing the cumulative total to 868
2026-05-14 16:42
Favorite

en.Wedoany.com Reported - The Cyberspace Administration of China (CAC) announced on May 13, 2026, that 72 new generative artificial intelligence services completed registration with the CAC between March and April 2026. During the same period, 49 new applications or functions that directly invoke the capabilities of already registered models via APIs or other methods completed filing. As of April 30, a cumulative total of 868 generative AI services had been registered nationwide, and 530 generative AI applications or functions had been filed.

The number of registrations continued the steady growth trend seen since the beginning of 2026. From January to February 2026, there were 48 new registrations and 46 new filings. As of February 28, the cumulative number of registrations nationwide was 796, with 481 filings. In the two months from late February to late April, the total number of registrations climbed from 796 to 868, a net increase of 72; the total number of filings rose from 481 to 530, a net increase of 49.

In terms of geographical distribution, the total number of registrations in multiple provinces reached new highs simultaneously. As of April 21, 2026, Shanghai had a total of 158 registered generative AI services, with the latest being the Tesla in-vehicle voice large model service, registered on April 20, 2026, by Tesla (Shanghai) Co., Ltd. As of April 23, 2026, Beijing had 225 large models registered with the CAC, accounting for about 30% of the national total and continuing to rank first nationwide. As of May 11, 2026, Guangdong Province had completed the filing of 53 generative AI services, with new models from companies such as Tencent Music, Gree Electric Appliances, China Southern Fund, and China Mobile Internet joining the list. During the same period, Jiangsu Province had a cumulative total of 74 registrations, with 7 newly registered services, including the "Big Cloud·Zhanlu R&D Large Model" from China Mobile (Suzhou) Software Technology Co., Ltd., and the "Aisi Large Language Model" from Nongxin Digital Intelligence Technology Co., Ltd.

The dual-track mechanism of registration and filing was reaffirmed in this announcement. The CAC clarified that those providing generative AI services with public opinion attributes or social mobilization capabilities can fulfill the registration or filing procedures through local cyberspace administration departments. Generative AI applications or functions that are already online should publicly display information about the registered or filed generative AI services they use in a prominent location or on the product details page, indicating the model name, registration number, or launch number. This mechanism is the core regulatory framework established by the "Interim Measures for the Management of Generative Artificial Intelligence Services," which was formulated based on laws including the "Cybersecurity Law of the People's Republic of China," the "Data Security Law of the People's Republic of China," the "Personal Information Protection Law of the People's Republic of China," and the "Science and Technology Progress Law of the People's Republic of China," implementing inclusive, prudent, and classified, tiered regulation for generative AI services.

Industry coverage continues to broaden. Judging from the recently published registration lists, registered services are no longer limited to internet companies and AI startups. Vertical large models from traditional industries such as manufacturing, finance, automotive, agriculture and animal husbandry, mobility, and retail are accelerating their inclusion. Guangdong Province's new filing list in May included enterprise-level applications such as Gree Electric's "Gree Intelligent Large Model," China Southern Fund's "China Southern Fund Financial Assistant Large Model," and China Mobile Internet's "Intelligent AI Response Large Model." In Jiangsu Province's new registration list, the "Kidswant Large Model" comes from the retail enterprise Kidswant Children's Products Co., Ltd., the "Ruige Aerospace Large Model" from Zhongke Ruige, and the "Qudoufeng Self-Driving Planning Large Model" from the travel and mobility sector. Nongxin Digital Intelligence's "Aisi Large Language Model" has been registered with the CAC, covering the agricultural field. These cross-industry cases indicate that AI large models are penetrating from general technology layers down into vertical industrial scenarios, and the registration system provides a unified entry point for compliant AI application across various industries.

Regulatory support is being advanced simultaneously. On May 8, 2026, the Cyberspace Administration of China deployed the "Qinglang·Rectify AI Application Chaos" special action, identifying the failure to fulfill large model registration and filing obligations as required, i.e., situations where registration is required but not completed, as one of the prominent issues for focused rectification in the first phase. On May 8, 2026, the CAC, the National Development and Reform Commission, and the Ministry of Industry and Information Technology jointly issued the "Implementation Opinions on the Standardized Application and Innovative Development of AI Agents," China's first normative document specifically targeting AI agents, making systematic arrangements for the governance framework, application scenarios, and industrial ecosystem of AI agents.

From the perspective of industrial governance, the registration system is evolving from a mere compliance threshold into a structural force driving industry standardization. The cumulative total of 868 registrations and 530 filings signifies that China's generative AI services have formed a multi-layered ecosystem covering both general large models and vertical applications. The filing system provides a lower compliance cost pathway for a large number of application-layer innovations that invoke registered models via APIs, allowing small and medium-sized enterprises and institutional users to embed AI capabilities into their own products without needing to repeat the full registration process. The balancing effect of this dual-track design, which aims to reconcile security and controllability with industrial vitality, is undergoing broader practical testing as the cumulative number of filings surpasses 500.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com