Subscribe Menu

Estonia’s Balancing Act Between Digital Innovation and Regulation, According to an Expert

We are living in a time of rapid technological growth and innovation. A key debate emerging from these developments is how best to balance innovation with regulation. Generally speaking, the European Union’s approach follows a more robust regulatory model. Meanwhile, the United States tends to favour a more entrepreneurial, hands-off approach.

Although Estonia has a strong pro-innovation culture, it must navigate this within regulatory frameworks. Eesti Elu spoke with Dr. Erkki Karo, a professor of science and technology policy at TalTech University, to learn how Estonia manages balancing regulation with innovation, while still emerging as a global leader in digitization and innovation. Find our conversation below.

Dr Erkki Karo
Dr Erkki Karo (source: taltech.ee)

Can you talk about your research interests?

My background combines governance and innovation studies. I study how we try to govern and manage technological change and development, and how governance systems influence how we approach technologies. At the same time, I look at how technologies, in turn, influence governance systems, both in Estonia and globally. Recently, I’ve been interested in topics related to sustainability and innovation.

How has Estonia adapted its governance structures to support digital innovation?

It’s kind of an interesting approach. If you think about how Estonia's public administration was set up in the early '90s—transitioning from Soviet legacies toward Western liberal democratic models—we opted for a very streamlined government structure. That aligned quite well with a certain model of digital governance.

Estonia has managed to do a lot in digital government over the past few decades because the governance structure was already quite simple. Technology has only reinforced that. Often, technology is used to inform existing policies, so it’s been a kind of back-and-forth evolution. The timing just fit well and created a path forward, both in terms of how we think about governance and how we approach digital technology.

“… our legal system is quite German in character. In those systems, you’re expected to legalize innovations before they happen, which is odd. You need to legalize experimentation for things to happen.”

(Dr. Erkki Karo)

But if you're more critical, you might say it's been a very linear approach where we've essentially digitized bureaucracy. Our governance system is still built on central state infrastructures—ID systems, registries, and so on—where the state remains a very central actor. Bureaucracy hasn’t disappeared; we’ve just hidden it. The simplified structure and individualistic policies have allowed that to happen.

One advantage is that capabilities have been built horizontally across the government. When new technologies emerge, there's no panic. We don’t need to immediately change the whole system. Instead, new tools get embedded across agencies. A few years ago, we had projects exploring AI in government, but there wasn’t a central plan. Different agencies experimented with the tech to see if it fit their work.

Now, AI is just another tool. If it gets modernized, we use it and try to make it fit. But we’re entering a phase where the tools themselves are transforming; digital technologies are evolving and our systems might not be adapting quickly enough. We assume old governance structures can integrate these tools, but they may not be able to use them to their full potential. So again, it’s a back-and-forth relationship.

How does the government balance regulation with innovation?

I don’t think we’re particularly good at regulation, partly because our legal system is quite German in character. In those systems, you’re expected to legalize innovations before they happen, which is odd. You need to legalize experimentation for things to happen.

Instead, what we tend to do is work through small networks, finding ways to reinterpret legal boundaries in light of new technologies. Learning first, then regulating and scaling them. For example, seven to eight years ago there were headlines about Estonia legalizing the piloting of self-driving cars. But we didn’t actually change any laws. We simply agreed to interpret existing laws differently. Estonian traffic law says that a driver must be licensed and sober, even if operating a car remotely—originally meant for things like towing. But we used that clause to allow remote-controlled testing via joystick. Once you’ve learned through such pilots, you can approach legal reform more confidently.

The broader challenge is that we have a pro-innovation culture and strong public-private collaboration, but rather than pass new laws, we tend to create shared understandings and then turn to formal legislation later.

How is Estonia building innovation policy for areas where there’s little precedent, like AI, and where experimentation is necessary?

It’s very decentralized. Ministries and agencies look at their own processes to see where a technology might apply. They each have their own space to test what’s feasible.

What institutional frameworks does Estonia have to promote innovation?

Historically, there’s been strong collaboration between the public and private sectors. Over the past twenty years, it’s been hard to say what’s public and what’s private in Estonia’s digital transformation.

But now the digital space is maturing, and we’re seeing more traditional industrial concerns emerge. So while we used to have strong public-private networks, those have started to fray in recent years. Still, we benefit from the decentralized model. We have pockets of efficiency. That helps maintain momentum, rather than relying on a single top-down agency to make all the decisions.

“Decentralization isn’t inherently more sustainable, but it does allow for trial and error. Instead of pivoting the entire system from one technology to another, we can test things in small pockets and scale what works.”

(Dr. Erkki Karo)

What are the effects of that decentralized model in terms of sustainability?

In our research, we’ve found that Estonia often assumes digitizing something automatically makes it more sustainable by reducing the human footprint. But we don’t really know if that’s true. We haven’t fully calculated the individual environmental impact of different digital technologies. The real question should be: what kind of digital government is the most sustainable?

Decentralization isn’t inherently more sustainable, but it does allow for trial and error. Instead of pivoting the entire system from one technology to another, we can test things in small pockets and scale what works. That makes the system more dynamic and hopefully more efficient in terms of learning and implementation.

How does Estonia collaborate internationally on innovation policy, including AI governance?

A lot of AI policy is driven back-and-forth with the EU, especially with the AI Act and efforts to set global standards.

But there’s a paradox. While Estonia is part of the EU, I think there’s often a preference for a more US-based approach, which is less regulatory and more innovation-friendly. In terms of broader innovation policy, you also have collaboration through the OECD and other organizations. But there are two stories in Estonia: in digital government, the state has led the way and been highly proactive. In other domains, it’s much more hands-off. Digital governance is unique in Estonia. It’s where the state is most interventionist. In other areas, the preference is often for the government to stay out of the way.

Are there any emerging trends in AI policy?

I’d say the key debate is between different models: the EU’s regulatory approach versus the US’s more innovation-driven one.

Estonia is caught between those two. We're part of the EU but we also want to use the best technologies, which often come from the US. AI development is very private-sector driven, and governments are always playing catch-up, deciding whether to remove barriers or impose new ones. So right now, the landscape is quite murky.

Responses have been edited for clarity and length.

This article was written by Natalie Jenkins as part of the Local Journalism Initiative.

Read more