Estonia develops digital identity system for AI agents
Summary and headline written by AI from the source article. How we work
The Estonian government is preparing a decision to allow artificial intelligence (AI) agents to securely perform daily digital tasks on behalf of individuals.
The government currently lacks a way to verify whether a digital service is being provided to a person directly or to their AI agent, and to determine the agent’s permissions and accountability. To address this, Estonia is drafting legislation and collaborating with the State Information Systems Department and service providers to prepare for initial testing.
According to Minister of Economic Affairs and Industry Liina Vahtras, Estonia aims to be among the first countries where AI agents can operate safely and under clear rules on a person’s behalf. The solution centers around a digital identifier, currently called “AI agent,” which will function as the digital identity of a user’s AI agent, allowing service providers to distinguish between actions taken by the person and those taken by their authorized agent.
An AI agent will not be a separate legal entity and will not have independent rights. All actions it takes will be linked to the person who initiated its use, and users will control the rights granted to their AI agent, such as limiting spending to 100 euros or restricting access to certain data. Service providers will also decide whether to allow AI agents to use their services and under what conditions.
Estonia intends to submit the draft law to the government soon and will begin testing once it is approved, with plans to seek recognition of the digital identity for AI agents at the European Union level.
