At the recent meeting of the ITU Telecommunication Standardization Sector (ITU-T) Study Group 17 in Geneva, Switzerland, a new international standard project titled "Trusted Requirements for Terminal Multi-Agent Systems," proposed by Ant Group in collaboration with the China Academy of Information and Communications Technology's Tieli Terminal Lab, China Telecom, and other organizations, was officially approved after full assembly discussion. This project, in response to the initiative "Building a Terminal Agent Ecosystem Together," has successfully been established as an international standard.

The standard focuses on four core dimensions: "trusted connection, trusted identity, trusted intent, and trusted authorization." It provides clear security guidance and technical basis for various stages of interaction among agents, including establishing connections, identifying identities, transmitting intentions, and executing authorizations. The goal is to establish a foundational framework for trusted interconnection, supporting the secure and healthy development of the industrial ecosystem.

Notably, the core technologies of this international standard originate from the Trusted Interconnection Working Group of IIFAA (Internet Trusted Authentication Alliance), which Ant Group participated in initiating. Previously, it promoted and open-sourced the industry's first intelligent agent trusted interconnection technology—ASL. This approval marks the formal entry of this open-source security collaboration framework into the international standard system, with its technical approach becoming an important cornerstone for building a global multi-agent trusted interconnection ecosystem.

image.png

(Figure caption: Ant Group-led international standard for trusted multi-agent systems successfully established at ITU)

The International Telecommunication Union (ITU) is a specialized agency of the United Nations responsible for information and communication technologies. Standards set by ITU-T are among the most authoritative international standards globally, alongside those from ISO and IEC. This approval indicates that the necessity and importance of the standard project have been widely recognized by international experts, reflecting the capability of Chinese technology companies to contribute technical practices in global digital governance.

With the rapid development of generative AI technology, agents have become key links connecting the physical and digital worlds. However, in cross-terminal and cross-platform collaboration, multiple agents face serious challenges such as malicious attacks, intent forgery, and privacy leaks due to the lack of unified security and trust standards. The newly approved standard aims to address this common challenge by systematically building a trusted architecture and full-process technical requirements for terminal multi-agent systems.

image.png

(Figure caption: Ant Group representative at the ITU-T SG17 meeting)

Deng Youjun, an expert from the China Academy of Information and Communications Technology's Tieli Terminal Lab, said: "Terminal agents are evolving from single-point intelligence to cross-terminal collaborative intelligence, and their trusted interconnection capabilities are a core indicator of digital infrastructure security. The initiation of this international standard fills the gap in the field of trusted cross-domain collaboration for agents, providing the industry with clear technical guidelines. It is expected to become a 'security pass' for high-quality industrial development."

The establishment of an international standard is a crucial milestone from 'zero to one.' In the future, the participating parties will conduct detailed drafting, discussion, and review over approximately two years within the ITU-T framework, aiming to make it an official international recommendation. The advancement of this standard will provide core support for building an open, trustworthy, and win-win global agent ecosystem.