UN telecom agency launches push for trust in AI agents

The International Telecommunication Union launched a new focus group on trust and identity for humans and agentic AI on July 9 at the AI for Good Global Summit in Geneva. The group will develop frameworks for trusted digital identity and preserve meaningful human control as AI agents take on tasks from scheduling and purchasing to more complex business processes.

Its description of agentic AI warns that these systems can act independently for users, impersonate people or organizations, and take unauthorized actions across interconnected digital systems. An agent can be asked to act, transact and carry out decisions in places where identity, permission and accountability matter.

ITU Secretary-General Doreen Bogdan-Martin said, “The future of AI depends on trust.” The focus group will be led by co-chairs Debora Comparin and Amir Banifatemi. The group will concentrate on trust management, agent identity and keeping humans in control while agents move across digital systems.

The body sits inside the ITU’s standards machinery rather than any enforcement arm. ITU-T Study Group 17 established it in June, and it reports into security standards work. Its influence will come through technical frameworks, interoperability expectations and guidance that governments and companies may later adopt.

AI for Good Global Summit 2026 ran July 7-10 at Palexpo, alongside the first UN-mandated Global Dialogue on AI Governance and the WSIS Forum 2026 as part of Digital Week. AI for Good was organized with more than 50 UN sister agencies and co-convened with the Government of Switzerland. On July 2, the ITU also launched the AI for Good Global Commission with more than 40 founding members, and 2.2 billion people still remain offline.

The ITU has set the focus group’s first meeting for Paris in November, followed by a second session in Geneva in January 2027.