This article is part 1 of the series “Public Sector 2025 – the technological shift shaping the future of public administration”, where we explore how AI, cybersecurity and modern software development together drive the next major leap in Swedish digital government.
AI in the Public Sector 2025: From Vision to Real Value
According to the Government’s Digitalisation Strategy 2025–2030, AI is now used to analyse large datasets, automate parts of case processing and support resource planning in healthcare. This enables faster decisions and a smoother experience for citizens. At the same time, the landscape is still fragmented. The solutions widely used today are primarily text- and data-driven, but several agencies are now testing more advanced multimodal systems that combine text, images and audio — mostly in pilots and innovation environments.
National guidelines for AI in the public sector came into force in 2025, reinforcing the focus on human oversight, traceability and clear documentation. This gives agencies a shared framework to rely on as the pace of adoption increases.
Sweden also builds from a strong foundation: around 27% of public organisations were already using AI in 2021, a figure that has continued to rise — most rapidly within central government.
At the same time as strategies and guidelines are being established, a growing number of practical applications are making a real difference in daily operations.
The Swedish Tax Agency: Smarter Flows and More Accurate Controls
The Tax Agency uses AI to categorise incoming emails, analyse large datasets and flag cases that require manual review. This frees up time for case officers and makes controls more accurate — both in preventing errors and detecting fraud early.
The Social Insurance Agency & Public Employment Service: Faster Cases with Higher Quality
Both agencies use AI to automate parts of case management within health insurance and labour market support. AI identifies anomalies, prioritises relevant cases and recommends actions — resulting in shorter processing times and increased legal certainty.
Healthcare & Social Care: Multimodal AI That Prevents Risk
Regions such as Uppsala are testing AI systems that analyse movement patterns of elderly individuals via camera. These systems can predict the risk of falls and alert staff in advance — a concrete example of multimodal AI where image data and motion analysis deliver direct societal benefit.
AI is also used in pilot projects to predict acute patient flows. By modelling demand at emergency departments, hospitals can plan staff and resources to avoid overcrowding and improve working conditions.
Climate & Sustainability: Collective Intelligence via RISE GPT
Twenty-five authorities and municipalities collaborate through the RISE GPT platform. AI is used to analyse energy consumption, environmental risks and urban planning — making it possible to compile facts faster and create individualised recommendations. The result is better decision support and more inclusive communication.
Three Success Factors for Scaling AI
As Swedish public organisations scale their AI initiatives, three success factors stand out clearly.
1. AI Leadership: Courage, Direction and Maturity
Leadership is the core of all public-sector AI development. All strategic frameworks point in the same direction: leaders must understand how AI reshapes workflows, competence needs and accountability. But just as important as broad organisational training is strong senior leadership support — the level at which priorities are set, resources are secured and transformation gains its mandate. Mature AI leadership requires more than understanding the technology. It means having the courage to make decisions in a rapidly evolving landscape, anchoring AI initiatives strategically, and creating stability through clear goals and governance. Senior leaders must actively support AI adoption, both in early strategic choices and throughout implementation:
- How do we integrate AI into core processes?
- What outcomes are we aiming for — and how do we follow up on them?
- How do we ensure human-in-the-loop and build trust at every stage?
More than 800 public-sector managers have already been trained in AI, ethics and trust-based leadership. These initiatives are essential — but the real impact emerges when training is combined with strong, senior-level support that enables organisations to scale AI from pilot environments into everyday operations.
2. Governance: The Structure That Makes AI Sustainable
AI governance has become a central pillar. In 2025, it is viewed as essential infrastructure — just as important as IT security and data protection.
Governance in practice revolves around three core areas:
- Accountability: Who owns the model and its decisions?
- Traceability: How can we see how the model reached its recommendation?
- Quality: How often are models tested and updated?
Shared AI workbenches, promoted by the Swedish Agency for Digital Government (DIGG), are highlighted as a national opportunity. With shared models, risk assessments and standards, the public sector can both accelerate adoption and strengthen security.
3. Societal Value: Clear Impact, Not Just Technology
Technology is the tool — not the goal. In climate, sustainability and citizen services, AI already contributes to faster analysis and more personalised information.
At the same time, the need for inclusive design is growing. Expert groups from AI Sweden and RISE emphasise collaboration, open data models and shared labs as key ways to reduce digital exclusion and create broad societal impact.
Where Are We Heading?
AI development continues to accelerate, and several areas are emerging as central going forward.
Digitala Twins for Urban Planning
By combining sensor, image and text data, municipalities can create digital twins of cities. This enables faster incident response, smarter traffic planning and better crisis preparedness. Several Swedish actors are already in pilot phases.
AI as Legal Decision Support
Automated tools that interpret legal texts, analyse decisions and help caseworkers navigate complex regulations may become the next major leap for legal certainty and efficiency.
Summary:
2025 is the year when AI becomes everyday reality — and when leadership, governance and societal value become the foundations that ensure development is both fast and safe.
Read part 2 of the series “Public Sector 2025 – the technological shift shaping the future of public administration”, focusing on Cybersecurity in the Public Sector 2025: offensive and defensive strategies.
Want to Put AI into Practice?
HiQ helps the public sector move from pilots to real value — with the right structure, accountability and technology to make AI work in day-to-day operations.
Get in touch — and let’s take the next step together!
