Estonia is considering issuing personal identification numbers to AI agents, separate from the people and organizations that deploy them. Prime Minister Kristen Michal said the plan would make Estonia the first country to assign such identifiers to AI assistants. The proposal aims to address a common practice today: when an AI agent acts on someone’s behalf, it often must use the owner’s identity and access, effectively operating “as you.”
According to reports from multiple outlets, the government’s approach is intended to clarify accountability when AI systems take actions for businesses, institutions, or individuals. By giving AI bots their own ID codes, the government would recognize them as distinct legal actors, potentially allowing legal rights to be associated with the agents and enabling responsibility to be attributed for actions taken under those identifiers.
The sources describe the policy as a backed proposal, but do not specify timing, technical implementation details, or how existing identity, authentication, and liability frameworks would be updated. The proposal is framed as a step toward more direct and traceable governance of AI behavior in real-world transactions and administrative processes.
