In a bid to foster both innovation and regulated application of generative AI services, cybersecurity authorities, in partnership with relevant government bodies, have been advancing the registration framework in accordance with the interim measures governing such technologies. Throughout the period spanning July to August 2026, a total of 124 new generative AI services secured official filing status with the national cyberspace administration. Among these newly recorded entries, seven are specifically tailored for on-device generative AI capabilities for mobile platforms.

For applications or functions that leverage already-filed model capabilities via API interfaces or other integration methods, a registration process conducted by local cybersecurity offices has been implemented. In this two-month window, an additional 133 such applications or features completed their registration. As of August 31, the cumulative tally shows 1,112 generative AI services having finalized their filing, while 731 generative AI applications or functions have successfully concluded their registration procedures.

The relevant details are now being publicly disclosed. Services that carry public opinion attributes or social mobilization functions are required to complete either the filing or registration process through their regional cybersecurity departments. Already-launched generative AI applications or functions are mandated to clearly display the status of the filed or registered generative AI services they employ, ensuring visibility in prominent locations or product detail pages. This disclosure must specify the model name along with either the filing number or the registration identifier.