Last week, at the Tallinn Digital Summit, we launched the second version of The Agentic State: a vision for how governments can use AI agents not to replace human judgment, but to redesign how the state itself works.
This new edition marks a significant leap forward, enriched by insights from digital leaders from around the globe and with The World Bank Group joining the Global Government Technology Centre as a supporting partner.
During the summit, I was asked to share what excites me most about the Agentic State. It’s not an easy question, but three reasons come immediately to mind:
1️⃣ It asks the fundamental question: Given current institutions and technological affordances, what must change so governments can deliver more for less? For developing countries, this is not a hypothetical question, it is an existential one.
2️⃣ It highlights a new technological inflection point: rapid advances in AI tooling, open-source models, and modular frameworks have drastically reduced the cost and time required to prototype, test, and deploy digital systems. Governments no longer need multimillion-dollar, multi-year IT projects to build usable solutions. If approached right, this shift creates new opportunities for problem-driven experimentation, potentially replacing the rigid, waterfall model that has long plagued public-sector technology delivery.
3️⃣ It opens a new path for low- and middle-income countries. As the paper notes, early government implementations show these capabilities translating to public administration contexts, with LMICs able to leapfrog more established digital government champions. This reframes the usual hierarchy of innovation: it’s not about catching up, but about building differently: lighter, faster, and closer to people’s needs.
The Agentic State reminds us that technology alone won’t transform government. Institutions, incentives, and imagination will. But with the right combination of all three, we can move from doing things right to doing the right things.
I’m grateful to have co-authored this with Luukas Ilves, Manuel Kilian, Simone Maria Parazzoli, Ott Velsberg, together with contributions from a global coalition of 21 digital leaders committed to ensuring AI serves the public good.