The new ITU Focus Group on Trust and Identity for Humans and Agent AI, announced at the AI for Good World Summit, comes at a time when artificial intelligence (AI) is rapidly evolving beyond assistance tools and increasingly becoming autonomous agents acting on behalf of people.
While agent AI promises significant productivity gains, it also introduces new risks, ranging from autonomous agents impersonating people or organizations to performing unauthorized actions in interconnected systems. The ITU Focus Group will address these challenges by developing frameworks that preserve meaningful human control in tasks such as executing financial transactions and operating critical infrastructure.
“The future of AI depends on trust,” said ITU Secretary-General Doreen Bogdan-Martin. “As AI becomes more autonomous, we must work together—industry, governments, academia, and civil society—to ensure the highest possible level of trust in AI systems.”.
As AI systems plan and act with increasing independence, the ability to establish an agent's identity and determine whether its behavior is trustworthy becomes critical.
Increasingly, AI agents need to identify and authenticate each other. Equally important, their decisions and actions must remain accountable, controllable, and reliable.
Identity systems establish who is acting, while reliability determines whether that actor is trustworthy. Together, they lay the foundation for safe interaction between humans and autonomous AI systems.
The Working Group will address the challenges of trust management for people and AI agents, the overall reliability of AI systems with agency capabilities, and ways to strengthen trust in the behavior of AI agents while maintaining authority over their actions.
“AI agents will soon be negotiating, transacting, and making decisions on our behalf,” said Working Group Co-Chair Debora Comparin. “Before that future becomes a reality, we need common international foundations that establish who these agents are, when they can be trusted, and how people will maintain control. That is the challenge this working group was created to address.”
“Agentive AI introduces a new class of digital actors who will increasingly collaborate with and among people,” noted Co-Chair Amir Banifatemi. “Identity tells us who is acting, and trustworthiness tells us how that actor can be expected to behave. Combining these two elements creates the common ground needed for interoperable, accountable, and reliable AI systems on a global scale.”
The group is open to technical experts, as well as specialists in policy, law, and regulation, in order to develop: The Working Group will report to the ITU expert group on security standards, ITU-T Study Group 17.
common terminology and definitions;
reference architectures for identity, trust, agent detection, and interoperability;
trust frameworks and lifecycle models (assurance);
interoperability mechanisms for digital identity and credentials;
security criteria and benchmarks for the ongoing assessment of AI agents; and
a standardization roadmap to coordinate action among expert communities.
“We are moving forward strategically and quickly, but deliberately taking the time to lay a solid foundation,” said Arnaud Taddei, Chair of Study Group 17. “We know the direction and, broadly speaking, what we want to build. Now we are bringing together the right leadership and structure, and ‘AI for Good’ is precisely where we will meet with partners who wish to join us.”.
The group's co-chairs will be joined by a broader leadership team, the composition of which will be finalized in the coming weeks following extensive consultation. The group, open to all interested experts, will hold its first meeting in Paris in November 2026 and its second in Geneva in January 2027.
ITU working groups help accelerate the development of new global standards and have been widely used to keep pace with the rapidly evolving AI landscape.
“The consensus we reached in developing ITU standards provides the clarity we need for new innovation ecosystems to grow,” said Seizo Onoe, Director of the ITU Telecommunication Standardization Bureau. “This working group is an open platform designed to help everyone move forward with confidence and certainty.”
Other active working groups address AI for smart cities, embedded and multimedia AI, and AI-native networks. Previous AI working groups have explored autonomous networks, health, agriculture, disaster risk reduction, and autonomous vehicles.
