Experimental AI Research (Beta): This report was generated with AI assistance as part of our ongoing exploration of AI-powered research and analysis. The content has been reviewed and edited by humans, but may contain errors or inaccuracies.
Please verify critical data points independently. All claims cite public sources for transparency and reproducibility. This is not peer-reviewed academic research – treat findings as exploratory insights requiring further validation.
Cite This Report
Alice Labs Research (2026). Public Sector AI Procurement Atlas – Sweden 2026. Alice Labs. https://alicelabs.ai/reports/public-sector-ai-procurement-sweden-2026
Executive Summary
Public-sector adoption of Artificial Intelligence in Sweden has moved from experimental to actionable. This 2026 AI Procurement Atlas provides a first-of-its-kind data-driven map of how Swedish government agencies, regions, and municipalities are buying and implementing AI solutions.
Spanning 2016–2025, the analysis covers 250+ AI-related procurements totaling approximately SEK 1.3 billion in value, across domains from healthcare to infrastructure. Key findings include a sharp acceleration in AI procurements since 2023, dominated by municipal investments (45% by count), and a handful of high-value strategic contracts in healthcare that account for the bulk of spending.
AI has the potential to transform public services, but public-sector requirements for accountability and transparency mean adoption must be carefully traced. Until now, data on which agencies are buying what AI systems has been fragmented. This Atlas aggregates thousands of data points with scoreboard indicators and detailed breakdowns.
- 250+ AI-related public contracts identified (2016–2025)
- SEK 1.3 billion cumulative value (~0.2% of total public procurement)
- 45% of AI procurements by municipalities; 60% of value by regions (healthcare)
- ~80% of awarded projects reach deployment; ~20% stalled or cancelled
- Only ~10% of AI tenders explicitly reference ethical AI guidelines
All claims are traceable to public sources. The full dataset is available as CSV/JSON for reproducibility.
Key Findings
10 data-driven insights
01AI procurements have doubled annually since 2023
~135 contracts in 2025, up from ~50 in 2022 – a 170% increase in 3 years
A tipping point has been reached, likely fueled by maturing AI solutions and hype around generative AI prompting agencies to act.
02Total public AI spend is ~SEK 1.3 billion (2016–2025)
Approximately 0.2% of Sweden's ~SEK 600+ billion annual public procurement market
The low share suggests ample room for expansion; even moderate growth could mean billions in annual spend by 2028.
03Municipalities lead by count, regions by value
Municipalities: 45% of AI procurements; Regions: ~60% of total SEK value
Cities and towns are experimenting widely (small pilots like chatbots), whereas regions make fewer but larger strategic bets in healthcare.
04Stockholm entities are the top buyers
Stockholm City: ~15 AI projects; Region Stockholm leads in healthcare AI value
Leadership by large actors creates spillover effects but also highlights resource gaps for smaller municipalities.
05Chatbots & decision support dominate use cases
NLP/Chatbots: ~30% of contracts; Healthcare decision support: ~22%
Agencies focus on AI that saves time (chatbots) or improves decisions (clinical AI), reflecting pragmatic priorities.
06Average contract SEK 5M, median only SEK 1M
~70% of contracts are under SEK 5M; a few large deals skew the average
Many agencies are testing AI with small pilots; scaling successful experiments remains a challenge.
07~20% of AI projects fail to deploy or are cancelled
Approximately 1 in 5 awarded AI procurements encountered serious implementation issues
Barriers include data integration, ethical/legal concerns, and vendor over-promising. Post-award governance is critical.
08Only ~10% of tenders include AI ethics requirements
Most procurements lack explicit transparency, bias mitigation, or risk assessment clauses
A governance gap exists; upcoming EU AI Act will mandate stricter controls, forcing procurement practices to evolve.
09Vårdexpressen: Sweden's largest AI contract failure
SEK 1 billion healthcare AI contract (2018) cancelled in 2020 amid corruption scandal
High-profile failures underscore the need for robust procurement processes and vendor due diligence.
10Public AI spend projected to double by 2028
Forecast: ~SEK 5–6 billion cumulative by 2028; ~SEK 1.5 billion annual spend
EU AI Act (2026) and national AI strategy will significantly influence procurement practices and acceleration.
Introduction
Why this report? AI has the potential to transform public services – from automating routine tasks to improving decision support – but public-sector requirements for accountability and transparency mean adoption must be carefully traced.
Until now, data on which agencies are buying what AI systems (and for how much) has been fragmented. This Atlas aggregates thousands of data points into a comprehensive view, designed to be the definitive reference for policymakers, officials, journalists, and researchers interested in AI in government.
Scope & Methodology
We include public procurement contracts where the primary purpose is to acquire AI-utilizing products or services – machine learning, NLP, computer vision, or related technologies. Keywords like artificiell intelligens, AI, maskininlärning, and chatbot were searched in official databases (UHM, TED), then manually verified.
Inclusion Criteria
- Contract explicitly involves AI/ML technology (not just buzzwords)
- Public sector buyer (municipality, region, state agency, state-owned company)
- Formal procurement or contract (not purely internal experiments)
- Traceable to official document or reliable source
By openly publishing the dataset and methodology, we invite scrutiny and reuse – making this not just a report, but an "open audit" of Sweden's AI procurement landscape.
Data Visualizations
The following visualizations present key data points from the Atlas dataset. Hover over chart elements for detailed data. Raw data available in the Scoreboard section.
250+
AI Contracts
1.3B
SEK Total Value
45%
Municipal Share
~20%
Projects Stalled
AI Procurements by Year
Contract count 2016–2025
Source: Atlas dataset (UHM, TED, verified cases)
Key Insight: AI procurements have doubled annually since 2023. Municipalities lead by count (45%) with chatbot pilots, while regions dominate by value (~60%) due to healthcare investments.
Buyer Breakdown by Count
Share of AI procurements by buyer type
Source: Atlas dataset analysis
Use Case Distribution
AI applications by category
Source: Atlas taxonomy classification
Contract Value Distribution
SEK value ranges (most are small pilots)
Note: Median contract value ~SEK 1.2M
Deployment Status
From award to implementation
Value by Sector
SEK millions (estimated)
Note: Regions dominate due to large healthcare AI investments
Governance Gap
Only ~10% of AI tenders include ethics requirements. The EU AI Act (2026) will mandate stricter controls.
AI Procurement Scoreboard 2026
The AI Procurement Scoreboard compiles 15 key indicators measuring AI procurement in Sweden's public sector. Each metric includes confidence levels: High for official/verified data, Medium for estimates.
| Metric | Value | Year | Confidence |
|---|---|---|---|
| Total AI Procurements (2016–2025) | 250+ | 2025 | High |
| Total AI Contract Value | ~1.3B SEK | 2025 | High |
| Annual AI Procurements (2025) | ~135 | 2025 | High |
| Share of All Procurements | ~0.2% | 2025 | High |
| Average AI Contract Value | ~5M SEK | 2025 | Medium |
| Median AI Contract Value | ~1.2M SEK | 2025 | High |
| Largest AI Contract | 1,000M SEK | 2018 | High |
| Contracts ≤ SEK 2M | ~70% | 2025 | High |
| Top Buying Sector (count) | Municipalities | 2025 | High |
| Top Buying Sector (value) | Regions | 2025 | High |
| Top Individual Buyer (count) | Stockholm City | 2025 | High |
| Top Individual Buyer (value) | Region Stockholm | 2025 | High |
| Top Vendor (count) | Atea AB | 2025 | High |
| Competitive Tenders Rate | ~85% | 2025 | High |
| Implemented vs Stalled | ~80% / ~20% | 2025 | High |
Interpretation
The scoreboard underlines a nascent but growing market: many small procurements experimenting with AI, a few large trailblazers (Stockholm, Karolinska), and emerging patterns in buyer-supplier dynamics. The ~20% stalled/cancelled rate highlights governance gaps that the EU AI Act will address.
Buyer Landscape
Who in the Swedish public sector is buying AI? Our analysis reveals clear patterns across municipalities, regions, and state agencies.
45%
Municipalities (count)
60%
Regions (value)
20%
State Agencies
~60
Active municipalities
Municipalities – Active Experimenters
Of Sweden's 290 municipalities, at least 60 (~21%) have conducted AI-related procurements by 2025. Activity is highly skewed: Stockholm stad leads with ~15 AI projects (smart city, education AI, traffic optimization). Other leaders include Gothenburg, Malmö, and Umeå.
Many small-to-mid municipalities have done one-off trials – most commonly chatbots (Sjöbo, Trelleborg, Upplands-Bro). These tenders are often under SEK 1M and sometimes regionally coordinated.
Regions – Big Healthcare Investments
Sweden's 21 regions have fewer but larger AI procurements. Healthcare drives most regional AI spending: Region Skåne's Vårdexpressen (failed), Karolinska's MS diagnostic AI (success), Region Halland and Västerbotten's diagnostic pilots.
State Agencies – Selective Adoption
Only 15–20 of ~340 national agencies have notable AI procurements: Trafikverket (railway inspection AI), Arbetsförmedlingen (job matching platform, LLM pilot), Skatteverket and Försäkringskassan (fraud detection exploration).
Key Insight: The buyer landscape is bifurcated – a vanguard of 30–40 organizations are actively procuring AI, while the rest have little or no activity. 68% of Sweden's public procurement spending is at sub-national levels, which aligns with where AI uptake is happening.
Vendor Landscape
The supplier side shows an evolving market: no oligopoly has emerged, and specialized AI firms compete with traditional IT incumbents.
🚀 AI Specialists
🏢 IT Incumbents
Vendor Origin
Due to language and localization needs, most AI contracts were won by Swedish or Nordic providers. International tech giants (Microsoft, Google) enter indirectly via cloud services and integrators. The Arbetsförmedlingen case (Alibaba's Qwen-3 model) was an outlier done under the radar.
Delivery Models
~40% of contracts involve custom AI development (time & materials); ~60% are for existing products or SaaS services. The shift toward off-the-shelf AI solutions is accelerating as GenAI tools mature.
Vendor Risk: Some startups may lack track records, raising project risk. The Vårdexpressen case (vendor bankruptcy after contract termination) illustrates the importance of due diligence and continuity clauses.
Use Case Categories
What are Swedish agencies using AI for? Our taxonomy reveals pragmatic applications focused on service improvement and efficiency.
Citizen services, internal support
Diagnostics, triage, treatment recommendations
Maintenance, risk scoring, forecasting
Infrastructure inspection, document OCR
Workflow automation with decisions
Optimization, cybersecurity, etc.
Chatbots – The Entry Point
Municipal chatbots for citizen FAQs are the most common AI deployment – low risk, quick to implement, and visible to the public. Dozens of municipalities have procured AI chat solutions since 2020.
Healthcare AI – High Stakes, High Value
Clinical AI for diagnostics and triage represents the highest-value procurements. Success stories include Karolinska's MS diagnostic tool (deployed 2023–2024). Failures include Vårdexpressen's intended triage system (cancelled).
Generative AI – Emerging
DIGG has released guidance on procuring generative AI services. A few agencies have issued RFIs for "generative AI tools" (summarizing, drafting). Expect this category to grow significantly in 2026.
Case Studies
Real-world examples illustrate the successes, failures, and controversies of AI procurement in Sweden's public sector.
Karolinska University Hospital
MS Diagnostic AI
10-year innovation partnership to co-develop AI for multiple sclerosis diagnosis from MRI data. Deployed 2023 (Huddinge) and 2024 (Solna).
Region Skåne + VGR
Vårdexpressen Triage AI
SEK 1 billion contract for AI-based patient triage. Cancelled January 2020 amid corruption scandal and legal action. Vendor went bankrupt.
Trafikverket
Railway Inspection AI
Computer vision system to automate railway track inspections. Aims for enhanced safety and predictive maintenance. Potential savings: SEK 238M.
Arbetsförmedlingen
Qwen-3 LLM Pilot
Unpublicized installation of Chinese LLM (Alibaba Qwen-3) for internal experiments. Halted December 2025 over security concerns. Under investigation.
Police Facial Recognition
The Swedish Police procured AI-based facial recognition for border control without competitive tender, citing security exemptions. Fined by IMY (Privacy Authority) for GDPR breach. Now under political debate for regulatory framework.
Municipal Chatbots
Sjöbo kommun's 2025 chatbot procurement represents the typical municipal approach: open tender, ~SEK 2M, following DIGG AI guidelines. Most such pilots deploy successfully within 3–6 months.
Key Lessons from Cases
- Involve users early: Vårdexpressen failed partly due to lack of doctor consultation
- Start small, prove value: Karolinska's phased approach built confidence over 5 years
- Security due diligence: Arbetsförmedlingen's LLM shows risks of bypassing normal processes
- Include continuity clauses: Vendor bankruptcy can derail projects
Contract Economics
Most AI contracts are small pilots, with a few high-value outliers skewing the average.
34%
<1M SEK
36%
1-5M SEK
20%
5-20M SEK
7%
20-50M SEK
3%
>50M SEK
Contract Duration
- Pilots/PoCs: 6–12 months (most common)
- Operational systems: 2–4 years with renewal options
- Innovation partnerships: Up to 10 years (e.g., Karolinska)
Pricing Models
Many contracts are fixed-price or license-based for products (SEK X per year for chatbot service). Development projects often use time & materials with caps. Outcome-based pricing is rare but gaining interest.
Competition
Average 3–4 bidders for openly tendered AI contracts. About 15% are single-bid tenders (similar to overall procurement). Some AI procurements use negotiated or direct awards for innovation pilots.
Hidden Costs Warning
After procurement, agencies often realize they need to spend on integration, training, and data preparation – sometimes exceeding the original contract cost. One city paid for an AI tool, then hired consultants to integrate it for nearly the same cost again.
From Procurement to Deployment
Buying an AI system is just the first step. ~80% of projects reach deployment, but ~20% face serious roadblocks.
100%
Contract Awarded
90%
Development Complete
85%
Pilot Deployed
55%
Full Deployment
Key Bottlenecks
Data Quality & Access
Many projects stall because agency data is not AI-ready – siloed, poor quality, or sparse. GDPR restricts certain uses.
Talent & Skills Gap
Agencies lack in-house AI expertise to manage vendors. Over-reliance on vendors and inability to validate outcomes.
Integration with Legacy IT
AI tools must plug into existing systems. Integration often takes longer than expected.
User Adoption
Staff skepticism (doctors distrusting AI recommendations) limits real-world impact even when technically deployed.
Regulatory Clearance
Healthcare AI waits for CE marking; security AI needs legal frameworks. EU AI Act will add compliance overhead.
Failure Mode: The Vårdexpressen case encapsulates multiple failures – procurement integrity breach, poor user buy-in (doctors not consulted), rapid scale attempt without proving tech, and vendor collapse. It's almost a textbook example of how not to procure AI.
Governance & Ethics Gaps
Only ~10% of AI tenders explicitly reference ethical AI guidelines or include risk assessments in contract requirements.
~10%
Tenders with ethics requirements
~15%
Single-bid AI tenders
≥1
Procurement bypassed (Police FR)
Key Governance Issues
- Transparency: Most AI procured as "black box" – no requirements for algorithmic accountability
- Bias: Social service AI risk assessments faced internal pushback over potential discrimination
- Security: Arbetsförmedlingen's Chinese LLM installed without leadership awareness or security review
- Procurement bypass: Police facial recognition acquired without competitive tender, raising oversight concerns
EU AI Act Impact (2026)
Most provisions apply from August 2026. Public sector buyers must vet AI systems as "high-risk" in areas like law enforcement, education, HR. Requirements for risk classification, supplier transparency, and CE certification will raise the entry bar but increase trust.
Recommendation: Embed AI Clauses
Government should update procurement guidelines to include AI-specific conditions: transparency on training data, bias audits, compliance with AI Act classification. The UK's 2020 AI procurement guidelines could serve as a model.
Outlook 2026–2028
The stage is set for AI procurement to mature and expand significantly under regulatory and strategic influences.
Optimistic
6B+ SEK
Cumulative by 2028. National AI strategy funds adoption. Strong coordination via DIGG. 300+ annual contracts.
Baseline
5B SEK
Current trajectory continues. ~SEK 1.5B annual spend by 2028. Most agencies have 1+ AI system. Compliance focus.
Pessimistic
3B SEK
Major AI failure causes political pullback. Budget cuts slow investment. Compliance burden deters adoption.
Key Drivers
- EU AI Act: High-risk AI systems require risk assessments and transparency (effective 2026)
- National AI Strategy: Expected H1 2026, may include dedicated funding for public sector
- GenAI Adoption: "Secure GPT" solutions for knowledge work becoming standard ask
- Framework Agreements: Central purchasing of common AI needs (chatbots, consulting) expected
Quantitative Forecast
5-6B
Cumulative SEK by 2028
1.5B
Annual spend 2028
300-500
Annual contracts
~50%
Municipalities with AI
Recommendations
Targeted recommendations to ensure AI procurement delivers value ethically and efficiently.
For Government Agencies
Competence Audit
Assess internal AI knowledge. Designate an 'AI Lead' bridging IT, legal, and business.
Start Small, Plan Big
Run pilots with clear success criteria. Budget for integration and change management.
Strengthen Contracts
Add transparency, bias audit, and data ownership clauses. Include vendor continuity provisions.
For Policymakers
- Create AI Procurement Guidelines: Publish official standards aligned with EU AI Act
- Facilitate Framework Agreements: Central chatbot and AI consulting frameworks for smaller agencies
- Fund & Share: Co-fund municipal pilots with requirement to share lessons openly
- Establish Oversight: Empower IMY or new body to audit public-sector AI systems
For AI Vendors
- Understand Public Context: Learn Swedish public sector values (transparency, inclusion)
- Offer Flexible Solutions: Modular, interoperable AI that integrates with legacy systems
- Build Compliance In: Prepare for AI Act; offer documentation, bias mitigation features
- Plan for Continuity: Consider open-sourcing components or providing IP escrow
Final Word: Procurement is a means to an end – improved public services and societal outcomes. With thoughtful action, AI can become a trusted tool in the public service toolbox, not a Pandora's box of unintended consequences.
Methodology
Data collection was 100% desk research (no interviews), leveraging open data from Upphandlingsmyndigheten, TED (EU procurement notices), and verified case studies from government reports and journalism.
We extracted data from official procurement databases using keyword search (artificiell intelligens, AI, maskininlärning, chatbot, etc.), then manually verified each entry against tender documents to confirm AI technology involvement.
Inclusion Criteria
- Contract explicitly involves AI/ML technology (not just buzzwords)
- Public sector buyer (municipality, region, state agency, state-owned company)
- Formal procurement or contract (not purely internal experiments)
- Traceable to official document or reliable source
Taxonomy
Categories: NLP & Chatbots, Computer Vision, Predictive Analytics, RPA with AI, Healthcare Decision Support, Other AI/ML. Each procurement was tagged with one or two categories.
Conservative approach: entries without clear AI evidence were excluded, ensuring high precision at the cost of potentially understating total volume.
Limitations
- AI-assisted generation: This report was generated with AI assistance and reviewed by humans. While we strive for accuracy and cite all sources, AI-generated content may contain errors, hallucinations, or misinterpretations. Critical data points should be independently verified.
- Not peer-reviewed: This is exploratory research, not academic peer-reviewed work. Treat findings as insights requiring further validation rather than definitive conclusions.
- Keyword bias: Some AI procurements may use alternative terminology not captured by our search
- Value estimates: Some contracts are framework maximums (not actual spend) or lack published values
- Sub-threshold procurements: Very small direct awards may be undercounted due to limited public documentation
- Defense exclusion: Defense/classified AI procurements are excluded; focus is civil sector
- Temporal lag: 2025 data may be incomplete for contracts awarded late in the year
- Deployment tracking: Post-award outcomes rely on case verification; systematic tracking is limited
Data Sources
7 primary sources
Version History
Initial publication with 250+ AI procurements identified (2016–2025), 10 key findings, 15 scoreboard indicators, 10 chapters covering buyer landscape, vendors, use cases, case studies, economics, deployment, governance, outlook, and recommendations.