Article content
What happens when artificial intelligence starts acting on our behalf? As AI agents become increasingly capable of handling everyday tasks, governments face a new challenge: how to embrace their potential without compromising trust, security and democratic values.
In the latest episode of The Art of Digitalisation, host Petra Holm speaks with Mihkel Solvak, Associate Professor of Technology Research at the University of Tartu, about the future of AI agents in digital government and what Estonia’s experience with digitalisation and internet voting can teach us about building trust in emerging technologies.
Can we trust AI agents?
Estonia has spent decades building a digital society where citizens routinely use online public services, digital identification and internet voting. According to Solvak, much of this trust has developed through experience: digital services have consistently delivered what people expect from them.
But will that trust automatically extend to AI agents?
The conversation explores why trust depends on how and where AI is used. While delegating routine administrative tasks, such as submitting annual reports, could save considerable time and resources, allowing AI to make consequential or irreversible decisions raises more difficult questions.
Solvak argues that governments should approach AI adoption gradually, experimenting with applications where the benefits are clear, risks are manageable and mistakes can be corrected.
More automation, but what about democracy?
One of the episode’s central questions is whether increasing automation could weaken citizens’ involvement in democratic processes.
Not every interaction with government should become effortless. While AI could eliminate unnecessary administrative work, democratic participation requires active engagement, debate and accountability.
The discussion also examines transparency and digital sovereignty. As governments become increasingly dependent on AI technologies developed by global technology companies, maintaining control over critical systems and preserving alternative ways to access public services become essential.
More freedom means more responsibility
Could AI agents make people less capable, or could they give us entirely new abilities?
Solvak offers an optimistic perspective. AI can empower individuals to accomplish things that were previously beyond their reach. However, greater freedom to delegate tasks also brings greater responsibility for the outcomes.
This creates new challenges for governments, from establishing clear accountability rules to helping citizens develop the skills needed to use AI safely and responsibly.
The episode ultimately raises a fundamental question: How can we build an AI-powered digital society that makes people’s lives easier without sacrificing human control, trust and democratic values?
Listen to the episode on Apple Podcasts and Spotify to explore:
- How Estonia’s experience with digital government and internet voting can inform trust in AI agents.
- Which government services could benefit most from AI-powered automation.
- Why democratic participation and human oversight must remain central.
- How AI agents challenge digital sovereignty, transparency and accountability.
- Whether AI will make people more dependent on technology or empower them to achieve more.
- Why governments should experiment carefully while preparing citizens for an increasingly AI-driven future.
As Estonia explores the next chapter of digital government, the conversation offers a timely reminder: technological progress is not only about what AI can do, but also about how societies choose to use it.
Listen to The Art of Digitalisation on Apple Podcasts and Spotify and watch for upcoming episodes.