The United Nations Development Programme (UNDP) has selected four municipalities – Skopje, North Macedonia; Rivne, Ukraine; Yerevan, Armenia; and Pristina, Kosovo[1] – for the first phase of the CEF x AI programme, following an application and assessment process.
The City Experiment Fund (CEF) works with municipalities across Europe and Central Asia to tackle complex urban challenges through strategic innovation and experimentation. It helps cities develop portfolios of connected actions that can be tested and adapted over time. CEF x AI extends this approach into the field of artificial intelligence (AI). AI is developing rapidly, while many municipalities are still building the policies, governance, skills and safeguards needed to guide its use in public services. The programme will help cities test, adopt, govern and benefit from AI in ways that are practical, inclusive and aligned with the public interest.
The programme runs along two complementary tracks:
- The CEF x AI Demonstrator supports municipalities that want to build longer-term institutional capacity to work with AI. Cities in this track will test and apply AI across different areas of city administration while strengthening internal systems, coordination, governance, and practical capabilities over time.
- The GenAI Challenge, delivered through UNDP’s BOOST impact acceleration programme, in partnership with IEEE Humanitarian Technologies, matches cities with external innovation teams to co-design open-source GenAI solutions for city-defined challenges within a shorter implementation period. The challenge brings IEEE’s network of engineers and technical experts together with UNDP’s experience working with municipalities, while city teams define the problems and help shape the solutions.
Together, the four cities’ work spans water and urban greenery, energy modernisation, air quality and transport planning.
- Skopje, North Macedonia: water and urban greenery – Skopje will explore how AI could strengthen the city’s stewardship of water and urban greenery through AI4Water and Green Cadastre, through the Demonstrator and BOOST tracks. These initiatives offer opportunities to learn how new evidence changes professional judgement, cooperation between municipal teams and public enterprises, and decisions about the city’s natural resources.
- Rivne, Ukraine: energy modernisation – Rivne will connect its wider energy-modernization agenda with a focused energy-efficiency calculator through the Demonstrator and BOOST tracks. The work will explore how AI could help homeowners, municipal teams and experts make informed choices about building improvements – and what changes in advice, coordination and municipal support are needed to turn those choices into action.
- Yerevan, Armenia: air quality and environmental governance – Yerevan will use air quality as a starting point for exploring AI’s role in environmental governance through the Demonstrator track. By bringing environmental evidence and residents’ perspectives into municipal decisions, the city will investigate how AI could support cooperation across departments, strengthen public participation and help translate information into more responsive environmental action.
- Pristina, Kosovo: transport planning – Pristina will explore how generative AI could support municipal transport planning through the BOOST track. A focused pilot will help planners bring relevant evidence together when reviewing transport routes and changing neighbourhood needs. The work will examine what makes AI useful and trustworthy in everyday planning, while keeping professional judgement and responsibility with municipal staff.
How the cities were selected
Ten applications were assessed by a selection panel comprising representatives of the Ministry of Finance of the Slovak Republic, the City of Bratislava, and UNDP. The panel reviewed them against published criteria covering strategic relevance, institutional ownership and delivery capacity, data and technical feasibility including responsible AI safeguards, implementation and sustainability, learning value, and fit with each track. Consultations with city teams followed before final recommendations were made.
What happens next
This autumn, the four cities will work with UNDP to turn their proposals into concrete plans and identify the first steps in each city. The programme aims to help cities shape responsible AI adoption around their own priorities. City teams and UNDP will test ideas, examine what changes in everyday municipal practice, and learn where AI creates public value and where its limits lie. What these four cities learn, including what does not work, will feed into wider regional learning about how cities can navigate AI-driven transformation in responsible way.
The programme is implemented by UNDP through the City Experiment Fund with the financial support of the Ministry of Finance of the Slovak Republic.
[1] References to Kosovo shall be understood to be in the context of United Nations Security Council resolution 1244 (1999).
