how estonia’s ai strategy is shaping the future

Estonia has successfully built a reputation as the world's most digital state, leading the way in e-governance, digital identity and almost 100% online public services. Now it wants to build a similar reputation with AI.

The Estonian government wants to see the productivity of workers doubled by 2035 using the technology, and has ambitions to give 100,000 people hands-on experience with AI skills within eighteen months. And while governments, businesses and citizens across the world wrestle with how to use AI responsibly and effectively, Estonia is already pushing ahead with a plan to introduce AI agent identities.

Here’s a look at Estonia’s ambitious future as a global AI leader.

Estonia's AI journey: from digital government to AI

Estonia kicked off its transformation into a digital leader back in 1996 with the launch of its Tiger Leap initiative. This was a country-wide IT infrastructure development programme, that helped lead to 99% of the population using the internet regularly, and put Estonia at the top spot in the Digital Development Index.

Since then, the journey has encompassed:

• the establishment of the first e-banking services in 1996
• e-Cabinet meetings, cutting government meetings from 5 hours to 30 minutes
• m-parking, which led to 95% of parking fees being paid via mobile phones
• e-ID and digital signature in 2002
• i-voting in 2005 – in the 2023 parliamentary elections history was made as 51% of votes were cast online
• World’s first e-⁠Residency programme in 2014

Out of all of these achievements, it’s e-⁠Residency that connects most closely with the plan to introduce AI agent identities. E-⁠Residency proved a country can offer trusted identity as a service across borders, and today more than 140,000 people from 187 countries hold an Estonian digital identity.

The Estonia AI strategy in action

Let’s examine the key pillars of Estonia’s AI strategy, which the country hopes will maintain its place as a global AI leader.

Eesti.ai

Estonia’s national AI initiative encompasses a number of key projects, with the overall aim of increasing GDP by 25% in five years and 50% in ten. Launched by Estonia’s prime minister Kristen Michal in January 2026, it is chaired by Bolt founder Markus Villig. 

In April 2026 its advisory board approved 15 flagship projects spanning digital skills, enterprise, the public sector, education, health and computing infrastructure – with the goal of bringing the practical use of artificial intelligence into everyday life.

AI Leap

In 2025, AI Leap was launched to bring AI into education. The successor to Tiger Leap, this initiative has already seen around 20,000 upper-secondary students and 3,000 teachers access leading AI-powered learning tools and training. 

It has already made Estonia one of the world’s first countries to introduce AI literacy nationwide, giving tens of thousands of youngsters a head start in using AI effectively and ethically.

AI-powered public services (Bürokratt) 

Estonia rolled out its AI-powered virtual assistant for public services in 2022. It's since been expanding gradually, with 18 public bodies live and more lined up to join, while its move to large language models is still being rolled out. It helps to manage enquiries, automates frequently asked questions, and if it gets stuck it will put a human into the loop.

Why AI agents need identities

One of the fears governments and the public have about AI is around accountability. This question has only become more important with the emergence of AI agents. These are essentially AI systems (usually run on an LLM) that can carry out tasks autonomously, such as reading and replying to emails, posting on social media, or using various types of software. This all raises the question – who is responsible for the actions of AI agents?

Estonia’s prime minister Kristen Michal has revealed plans for the country to be the first in the world to introduce AI agent identities through registered agent codes. With this ID code, an AI agent would be trackable and accountable for its actions, with a visible chain of responsibility. This means it would be possible to see which agent acted, who authorised those actions, and what limits were set for the agent. Ultimate responsibility would therefore remain with the person or company behind the agent. 

Other countries’ initiatives to make AI accountable

Estonia is not alone in trying to solve the accountability problem. This table offers a quick comparison of various countries’ AI accountability initiatives and where they are in their development.

What Estonia's AI future means for entrepreneurs and e-⁠residents

While Estonia has big plans for AI in the future, it's already part of the country's business and bureaucratic environment today. Many e-⁠residents already use generative AI for everything from content and admin to research, running lean, AI-native businesses without waiting for a formal framework to catch up.

The bigger shift lies ahead. If AI agent identities become available, routine company admin, filing, invoicing, even VAT declarations, could all eventually be handed to an agent acting within limits the founder sets. The founder would keep final control and a clear record of what the agent did. This isn't live yet, and the advisory board has only agreed to move ahead with the concept. E-residents should keep an eye out for when that might be rolled out. Subscribe to the e-⁠Residency newsletter to stay informed. 

What's more useful than the timeline is the direction it points to. Any future AI ID code would likely build on the same trust and identity infrastructure e-⁠Residency already runs on, rather than a separate system bolted on top. That matters for founders who are deciding where to start their company, as it reinforces Estonia as a credible base for AI-driven businesses. And whatever agents eventually take on, the principle stays the same as everywhere else in this plan: the agent doesn't get independent authority, the founder does.

FAQs

What is an AI agent?

An AI agent is a piece of software that can carry out tasks by itself. It doesn't just answer a question and stop. It can read an email, write a reply, and send it. It can fill in a form, make a booking, or handle a task start to finish. A person sets the goal for the agent, and the agent does the work automatically.

Which country is number one in artificial intelligence?

The United States leads most global AI rankings, including the Tortoise Global AI Index and the Stanford AI Index. China comes second. These rankings measure things like research output, investment, and talent. Estonia isn't chasing that title. Its focus is on using AI well inside government, in ways people actually experience day to day.

Why is Estonia so digitally advanced?

After it regained independence in 1991, Estonia invested early and heavily in internet access and digital ID, then kept building on top, including online banking, e-voting, and digital signatures. Three decades on, almost all public services run online. AI is simply the next layer on that same foundation.