The Complete Guide to Using AI in the Government Industry in Slovenia in 2025

Slovenia's government moved from policy to practical rollout in 2025, building on the National Programme to Promote the Development and Use of Artificial Intelligence by 2025 (NpAI) to bring AI into public administration, boost research capacity and expand training for civil servants; the programme's human‑centric goals and a planned regulatory sandbox sit alongside national steps to transpose the EU AI Act and designate supervisory authorities as part of broader implementation work documented by the government (National Programme to Promote the Development and Use of AI in Slovenia - National Programme for AI 2025, Act on implementation of the EU AI Act - Government of Slovenia).

Strong research assets - notably the Vega supercomputer and the UNESCO‑affiliated IRCAI - underpin pilots and public‑sector use cases, while workforce gaps make practical training essential; for public servants and managers wanting job‑focused AI skills, the AI Essentials for Work bootcamp offers a 15‑week, workplace‑centred syllabus and a clear registration path (AI Essentials for Work bootcamp registration - Nucamp).

“We're not trying to compete with Silicon Valley - but we can show how AI can be democratic, useful, and locally accountable.” - Dr. Marko Grobelnik, Jožef Stefan Institute

The National Programme to Promote the Development and Use of Artificial Intelligence (NpUI or NpAI) lays out a practical, multidisciplinary road map for Slovenia through 2025: it combines funding, infrastructure and human‑capital measures to turn research into real public‑sector and industrial solutions, earmarking EUR 110 million of public funding to implement these strategic objectives until 2025; the plan explicitly targets creating a supportive ecosystem, strengthening industrial and technological capacity, rolling out reference AI solutions in healthcare, Industry 4.0, culture and public administration, and setting up a National AI Observatory alongside a strong ethical and legal framework to build public trust.

Education and lifelong learning are core pillars - from adding AI modules in primary and secondary curricula to web‑based upskilling for professionals - while lab‑to‑market measures include co‑financing for research across TRLs and support for standardisation and international cooperation.

For a concise breakdown see the European Commission Slovenia AI Strategy report and the full NpAI document on DIGI.WATCH, which together show how data and HPC resources (like Vega) and institutions such as the IRCAI are woven into the strategy to move AI from pilots into everyday public services and industry.

Slovenia's AI governance is being pulled together around the Ministry of Digital Transformation, which is not only shepherding national strategy and hosting one of the EU's new AI factories with a high‑performance supercomputer, but is also preparing the next strategic framework and inviting public input on AI policy (Slovenia Ministry of Digital Transformation - official site; see the recent public consultation announcement).

The Ministry has been formally positioned as the primary authority to implement the EU AI Act and is supported by an expert council to advise on transposition, oversight and innovation-friendly tools such as regulatory sandboxes, while proposals under consideration even include an AI Ethics Commissioner to boost trust and accountability (Slovenia AI Act national implementation and governance measures).

Practical buy‑in on the ground is encouraged via public comment processes - stakeholders were invited to respond to the 2026–2030 strategic framework - so ministries, standards bodies and named advisors like senior policy experts play active roles in turning strategy into funded projects and safe pilots (Slovenia public consultation on AI policy 2026–2030).

The result is a compact governance ecosystem: a central ministry driving coordination, specialist advisors guiding standards and ethics, and clear lines for sandboxing and supervision so that infrastructure investments (the AI Factory/supercomputer) can translate into accountable public‑sector AI services.

“A major step towards a better digital future for our country” - Minister Klampfer

Slovenia moved from strategy to concrete rules in 2025: at the 21 August government session it adopted a draft Act to implement the EU AI Act that explicitly maps out supervisory authorities, names the competent body for regulatory sandboxes and even introduces the possibility of appointing an AI Ethics Commissioner - steps that the government says place Slovenia among the leading EU Member States for comprehensive AI regulation (Slovenia government 164th session: draft Act to implement EU AI Act).

Those national moves sit inside the EU's risk‑based architecture - the AI Act entered into force on 1 Aug 2024, with prohibitions and early literacy duties taking effect from Feb 2025, GPAI governance rules from Aug 2025 and full applicability scheduled for 2 Aug 2026 - meaning ministries and deployers must prepare for strict obligations (risk assessments, logging, documentation and human oversight) for high‑risk systems and transparency duties for limited‑risk tools (EU AI Act regulatory framework - European Commission).

The practical upshot for public services is clear: designated authorities, sandbox pathways and defined penalties will make safe experimentation possible, while the looming EU timelines turn planning into an urgent operational checklist - imagine an ethics commissioner as a referee in the stadium of public AI, blowing the whistle when systems threaten rights.

Data and compute are the quiet engines behind Slovenia's AI ambitions: at the heart of public data is OPSI, the national open‑data portal built on the CKAN platform that catalogs datasets from ministries, municipalities and agencies and even feeds metadata to the European Data Portal so developers across Europe can build on Slovene public information; OPSI's built‑in publishing workflow, Drupal content management and secure CAS login mean datasets are vetted, draftable and published with machine‑readable metadata, while automatic checks that flag broken links and a

“dataset openness” score

act like a quality‑control referee keeping the catalogue usable for pilots and services.

That practical polish matters when linking open data to heavier infrastructure - national data spaces and HPC resources (already cited in earlier sections such as the Vega supercomputer) need clean, licensed, machine‑friendly feeds to power OCR, spatial planning and health analytics without hidden surprises.

For anyone planning public‑sector AI projects, OPSI is the single, well‑documented entry point for reusable data and APIs, and Slovenia's portal design (CKAN + editorial workflow + international harvesting) shows how open data can be both a public good and a reliable building block for production AI systems.

Slovenia is building human capital for public‑sector AI by knitting classroom reform, teacher development and adult upskilling into a single, practical pipeline: the national Digital Education Action Plan (ANDI 2021–27) and projects like the eTorba e‑textbook platform aim to bring computing, AI basics and gigabit connectivity into everyday schooling, while targeted initiatives such as the 2023–26 “Digital and sustainable teacher” programme and pilots like B‑RIN and Innovative Pedagogy 5.0 focus on teacher capacity and didactic methods for computing and informatics; at the same time the 2022 Act on Promotion of Digital Inclusion backs access with measures including a digital voucher (€100–200) to buy devices or training.

That mix matters because recent research flags clear gaps: an IEEE study found Slovenian university students lagging on practical AI knowledge - for one question about where AI is typically applied they scored only 7.8% - underscoring why lifelong learning and workplace‑centred reskilling must run alongside curriculum change.

For officials planning rollouts, the lesson is simple and vivid: clean infrastructure and open data (OPSI, Vega) are necessary, but without teachers and civil servants fluent in AI concepts and ethics, pilots won't scale into trustworthy public services (Slovenia Digital Education Action Plan (ANDI) and eTorba e‑textbook platform, IEEE Access AI literacy study on Slovenian university students, Slovenia education technology profile - digital inclusion and policy).

Slovenia's priority AI use cases in the public sector cluster around health, crisis analytics and cultural‑language services: health and medicine sit at the top of the list, with certified clinical decision support systems such as SBAS and consumer tools like mySmartBlood that interpret blood parameters, age and sex to surface the most likely conditions and aid faster diagnosis (Smart Blood Analytics - SBAS & mySmartBlood clinical decision support), while the country's mature eHealth backbone (openEHR-based national EHR and near‑universal ePrescription use) makes those AI tools immediately relevant to clinical workflows and scaling; in fact, ePrescription adoption is at 99.5% - virtually no paper prescriptions remain (Slovenia eHealth adoption and openEHR overview).

Reference projects also include pandemic and public‑health modelling from the Jožef Stefan Institute that blend SEIR models with machine learning for workload and infection forecasts, showing how AI supports operational decisions in crises, and smaller domestic innovators and integrators are building applied solutions (telemedicine, medication management, predictive analytics) ready for pilots.

Finally, cultural‑heritage and language technology use cases - for example Slovene‑language OCR pipelines to unlock archives and multilingual services - are highlighted as practical, lower‑risk ways for ministries to deliver public value while building data assets and skills (Slovene‑language OCR pipeline for cultural heritage - Nucamp); the combination of certified medical algorithms, a near‑complete digital prescription system and active research labs makes Slovenia's public‑sector AI agenda unusually well‑positioned to move from pilots to production without reinventing the plumbing.

“By today, 99.5% of providers are using ePrescription... to connect to the national EHR. In case of lack of compliance, the manager of the healthcare provider will get a fine.” - Alenka Kolar, Acting Director‑General of the Directorate for Digitalisation in Healthcare, Ministry of Health Slovenia

Ethics, oversight and public trust are front and centre as Slovenia moves from plans to practice: the government's 21 August 2025 draft Act to implement the EU AI Act formally maps supervisory authorities, names the competent body for regulatory sandboxes and even opens the door to appointing an AI Ethics Commissioner (Slovenia draft Act to implement the EU AI Act - government 164th session (21 August 2025)), while the National Programme and the European Commission's AI Watch highlight parallel work to create an ethical framework, a national supervisory mechanism and a National AI Observatory to monitor uptake and build trust (European Commission AI Watch: Slovenia AI strategy report).

That institutional scaffolding is being matched by international norm-setting and industry buy‑in after the Brdo forum, where global firms pledged to follow UNESCO's ethics recommendations - a practical signal that private‑sector practices will be watched alongside public rules (Tech giants commit to UNESCO ethical AI framework at Brdo forum).

Practically, this means sandboxes, transparency duties and outreach (AI ambassadors, conferences and media dialogue) are not optional extras but tools to prevent fast, invisible harms such as algorithmic bias or AI‑driven disinformation; in a world where a single viral deepfake can undermine confidence overnight, Slovenia's blended approach - law, ethics oversight, public engagement and international alignment - aims to make accountability palpable to citizens and auditable to regulators.

“This alliance of the public and private sectors is critical to building AI for the common good.” - Audrey Azoulay, UNESCO

Which country has the highest use of AI? On metrics that measure active learning and developer engagement, the United States sits well ahead - ApX's 2025 AI Engagement Rankings put the U.S. at 100.00 with China, India and Germany trailing much lower - while Stanford HAI's 2025 AI Index underscores the same reality from the supply side (the U.S. still produces the most notable models and saw massive private investment in 2024), so leadership today is as much about people learning and building as it is about capital and compute (ApX 2025 AI Engagement Index country rankings, Stanford HAI 2025 AI Index Report).

Slovenia's position in the engagement ranking (around 0.51, rank ~71) shows a clear gap between national assets - like a humming Vega supercomputer and open OPSI data - and the active learner base needed to put those resources to work; in plain terms, Slovenia has powerful engines parked in the garage until more developers, civil‑servant reskilling and demand for production AI are mobilised.

That gap points to a practical roadmap for public agencies: convert infrastructure and strategy into hands‑on training, industry partnerships and citizen‑facing pilots so Slovenia's technical promise translates into widespread, daily AI use.

Practical next steps for putting Slovenia's AI strategy into everyday government work are straightforward: focus first on people and practice - scale staff training and lifelong learning so ministries can turn the EUR 110 million NpUI commitment and world‑class compute (Vega) into repeatable services rather than one‑off pilots; use regulatory sandboxes and “learning by doing” pilots to validate reference solutions in health, public administration and language tech; publish deployments and foster transparency so citizens can monitor AI in action; and build or fund a national Digital Innovation Hub to broker data, private‑sector access and cross‑institution projects.

The European Commission's AI Watch synthesis lays out these same building blocks - training, sandboxes, open data and observatories - as the public‑sector pathway from lab to production (European Commission AI Watch - Slovenia public-sector AI strategy), while the Ministry of Digital Transformation is already positioning Slovenia's AI Factory and HPC as practical infrastructure to support pilots (Slovenia Ministry of Digital Transformation - AI Factory & Vega).

For teams that need immediate, job‑focused skills to run and govern these pilots, the AI Essentials for Work bootcamp offers a 15‑week, workplace‑centred curriculum and a clear registration route to get civil servants and project managers hands‑on quickly (AI Essentials for Work bootcamp - 15-week workplace AI training (register)); the “so what” is simple: without training and transparency, great compute and open data stay powerful but idle - train people, trial responsibly, publish results, repeat.

“A major step towards a better digital future for our country.” - Minister Klampfer, Ministry of Digital Transformation