Skip to main content



AI and the Kratt* momentum

In 20 years, Estonia has built a society where public and private digital services are woven into the fabric of everyday life. Armed with secure digital identities, citizens can conduct their governmental and private business online. Marriage, divorce, and real-estate deals are among the rare exceptions. You still need to leave the house for those.

A digital society only works when people actually trust and use digital channels. By and large, Estonians do. This pragmatism has helped the country gather mountains of anonymous raw data. This is fuel for machine learning and other technologies.

The Estonian government now plans to start building next-generation public services based on artificial intelligence, or AI. This creates interesting obstacles. The government’s newly formed AI taskforce aims to solve a few critical ones. Among them: legal liability. Who is responsible when artificial intelligence messes up?

Algorithmic identity

A year into the public debate over algorithmic-liability law (a.k.a. the Kratt* law), opinion leans toward avoiding sector-based regulation, opting for general algorithmic liability instead. The biggest conversation starter is the idea of giving algorithms a separate legal status, similar to companies. This is one of the more future-proof methods of keeping AI innovation humming in the longer term. A draft bill should enter parliament for debates by June 2019.

Aside from the ethical, moral and philosophical debate over AI liability, the task force is also working on enforcement. How does one ensure the integrity of decision-making within algorithms that evolve and change constantly? How can we be sure that sensor data used in algorithms has not been hacked or changed? Estonia’s experience in using blockchain technology to secure its citizens’ medical records may help.

A balanced debate on AI

In these debates, technical and legal expertise goes a long way. But the discussion must also involve the public. Honest, meaningful debate requires that dreamy utopias be balanced with open discussions about AI’s controversial attributes and threats. Only this can create user-friendly legislation that’s equipped to reduce legal nightmares in the long-term.

The AI taskforce has another mission. It is working on a national AI strategy to create a common framework for implementation in both the public and private sectors. Estonia is a frequent testbed for new technologies and sees AI as just the latest entry. If it can reduce regulatory complexities in a user-friendly way, the path is clear for mind-blowing innovation.

Our future with AI in Tallinn

Naturally, Estonia sees itself fit to lead the global debate on AI. On the 16th of October 2018, it will host the second Tallinn Digital Summit, a platform for discussing the global future with self-learning algorithms and their impact on social systems.

Held annually, it’s an invite-only gathering of heads of government from digital-minded nations, the tech community, and influential thinkers. This year’s summit will focus on putting AI and free-moving data in the service of digital governments, economies and societies.


* Kratt is a magical creature in Estonian mythology. Essentially, Kratt was a servant built from hay or old household items. The Estonian government uses this character as a metaphor for AI and its complexities. #krattlaw, #ailaw, #algorithmicliability

read more on similar topic

visit the e-estonia briefing centre

The goal of e-Estonia Briefing Centre is to inspire global policy makers, political leaders, corporate executives, investors and international media with the success story of e-Estonia and build links to leading IT service providers.

The e-Estonia Briefing Centre has become a must-see destination, hosting over 10,000 international decision-makers every year. Make sure to book your visit in advance.

Give feedback