Estonia plans to be the first nation to grant official digital identities to artificial intelligence agents. This initiative extends the country’s long-standing digital governance model. Prime Minister Kristen Michal announced the plan, stating it would allow software to act on behalf of individuals or corporations without requiring full access to the human user’s credentials, rights, and data.
The proposal, backed by the Eesti.ai advisory board, aims to create an “AI ID code.” This digital identity would enable agents to perform specific digital tasks, such as compiling reports, drafting declarations, or interacting with information systems. The core concept is delegating authority that is limited, controllable, revocable, and auditable. Instead of granting broad access, the AI ID code would specify narrow permissions—whether the agent can only view records, draft documents, or authorize a specific payment.
Estonia has a proven digital infrastructure, having implemented personal identification codes and the e-ID program since 2002, alongside the X-Road data-exchange backbone. Michal emphasized that the success of Estonia’s digital state was built on trust. By establishing this official identity, Estonia aims to become the first to govern AI agents formally.
While the plan details the technical structure, it leaves the crucial question of liability unresolved: who is responsible if an agent with its own ID causes an error? Currently, liability typically rests with the deploying entity. The development of an agent-specific identity challenges this default structure.
Internationally, other jurisdictions and private companies are exploring similar frameworks, highlighting the global nature of the emerging need to govern automated actors.
Topics: #first #estonia #give
