Estonia launches first AI digital identities to let agents act on behalf of people and companies with defined limits

Estonia will become the first country to issue digital identity codes for artificial intelligence agents, the government said. The new system lets companies and individuals deploy AI agents to perform tasks on their behalf without granting those agents unrestricted access to personal data or services.

"In the future, AI will increasingly carry out digital tasks on our behalf, compiling reports, preparing declarations or interacting with information systems," Estonian Prime Minister Kristen Michal said in a statement. "To that end, it must be clear who is acting on whose behalf with what rights, and who is ultimately responsible."

The prime minister's office said the AI ID codes will prevent situations where individuals or organisations must give AI assistants access to all their rights, services and data. Agents are systems that perceive information, make decisions and autonomously perform tasks to achieve specific goals. Major companies including OpenAI, Anthropic, Google and Microsoft already deploy agents inside their popular chatbots.

Why traditional authentication falls short

Estonia's move aligns with recent research showing that conventional identity frameworks cannot govern agents that act, decide and transact at machine speed. Multi-factor authentication, the standard for logging into banking apps, was designed for humans who pause between steps. An AI agent operating autonomously needs a different model - one that defines its scope and accountability before it starts working.

The Baltic country ranked fifth among the top 10 most digitally progressive states in the Organisation for Economic Co-operation and Development's 2026 assessment, tied with Norway, Ireland and Denmark. Estonia already runs a state-backed e-ID system that gives citizens access to public services online. Its m-Residency program lets foreign nationals operate digital-first businesses using the same mobile ID issued to citizens, from anywhere in the world.

Building on existing digital infrastructure

The AI identity layer extends Estonia's existing digital government architecture rather than replacing it. Citizens and organisations will assign the codes to AI agents with defined limits, creating a clear chain of responsibility. The government's statement emphasised that ultimate accountability stays with the human or organisation that deploys the agent.

For policy makers tracking how governments regulate autonomous systems, this development offers a concrete case study. Estonia is codifying the principle that an agent's permissions must be scoped and auditable, not open-ended. The approach mirrors how the country already treats digital identity for humans - as a controlled, verifiable credential rather than a blanket login.

Why this matters for government professionals

Estonia's AI ID framework sets a precedent that other governments will study closely. The core question is not whether AI agents will access public services, but under what rules. For civil servants and policy makers, the announcement signals that identity infrastructure must evolve from authenticating people to authenticating delegated authority. The technical and legal standards decided now will shape how agencies procure AI tools, audit automated decisions and assign liability when an agent acts outside its scope.

Professionals working in AI for Government roles or following an AI Learning Path for Policy Makers will find Estonia's implementation worth watching as it moves from announcement to live deployment.