2026-09-18 AI Agent Identity News
Summary: Daily digest for 18 September 2026. No new Estonia development. The ITU, with World Bank Group participation, launches a new international focus group (TIDA) on trust and identity for humans and agentic AI, framing agent identity as a delegation trust chain rather than a legal identity of the agent itself.
Sources: 2026-09-18-ai-agent-identity-news.md
Last updated: 2026-09-18
Section 1 — Estonia
No new non-duplicate sources materially advancing Estonia’s AI-agent identity, tehisaru / isikukood, delegated-authority, authorization, or accountability discussion were found in this coverage window (source: 2026-09-18-ai-agent-identity-news.md). The 2026-09-08 correction — no personal identification codes for AI agents, responsibility stays with the human — stands unchanged. See estonia-ai-agent-id.
Section 2 — Other countries / international
ITU and World Bank launch Focus Group on Trust and Identity for Humans and Agentic AI (TIDA)
The International Telecommunication Union announced a new Focus Group on Trust and Identity for Humans and Agentic AI (TIDA), with World Bank Group participation, published 17 September 2026 (primary international standards/digital-development source). The group’s planned scope covers terminology, use cases, identity models, credential formats, assurance levels, lifecycle controls, and mechanisms for government systems to verify an agent’s authorization before it acts (source: 2026-09-18-ai-agent-identity-news.md, citing ITU, 2026-09-17).
Notably, the ITU frames an agent credential as evidence linking the agent to the delegating person or organization, specifying permitted capabilities and mandate duration — calling for delegation that is specific, time-bound, traceable, and revocable, with auditable evidence per action. This treats agent identity as a trust chain (person/organization → delegation → agent → action) rather than a standalone legal identity for the agent, and aims at mutual recognition of agent authorization across agencies and borders. See agent-standards-and-interoperability and agent-authorization-and-delegation.
No other new country-level policy, government, regulatory, standards, or major-news source was found after deduplication (source: 2026-09-18-ai-agent-identity-news.md).
Section 3 — Research
Not applicable. This report runs on a Friday; the Monday research section is intentionally omitted (source: 2026-09-18-ai-agent-identity-news.md).