ITU stands up a Focus Group on trust and identity for AI agents

The ITU has created a Focus Group on Trust and Identity for Humans and Agentic AI, chartered to build the terminology, reference architectures and trust frameworks for establishing who an AI agent is and whether its behavior can be trusted. Announced at the AI for Good Global Summit and reporting to ITU-T Study Group 17, it is tasked with agent-discovery and interoperability specs, lifecycle assurance models and continuous security benchmarks, with a first meeting in Paris in November and a second in Geneva in January (via ID Tech Wire).

This is the standards machinery of digital identity being pointed at software that acts on your behalf — the same question NEC and Ava Labs are chasing from the vendor side (also in today’s notes). “Identity systems establish who is acting, while trustworthiness determines whether that actor is reliable,” as Secretary-General Doreen Bogdan-Martin frames it, and both are wanted before agents execute transactions or touch critical infrastructure. A UN body chartering agent-identity work is an early sign that the topic is moving from conference panels to formal spec.

Focus Groups are where the ITU floats an area before committing a study group, so this is early and non-binding. But terminology and reference architectures are exactly where the durable definitions get set. Worth tracking whose architecture wins.