🔍 Read the full analysis: What AI Adoption Looks Like Across Europe’s Governments on ThorstenMeyerAI.com
Get smart everyday buys delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Developments reported across the Netherlands, Britain and Sweden show governments using AI in staffing plans, asylum casework and public administration, while also restricting some tools and confronting bot-generated consultation responses. The figures and policies differ, and they do not establish that AI alone is driving job cuts or government decisions.
Recent developments in the Netherlands, Britain and Sweden show European governments taking sharply different approaches to artificial intelligence: the Dutch government is planning a smaller civil service with AI among several proposed tools, the UK Home Office uses AI to support asylum casework, and Sweden has blocked public versions of ChatGPT and Claude on staff computers. A separate UK review found that bots accounted for 76% of responses to a 2023 health consultation, highlighting a challenge for public participation as well as government operations.
The Dutch government’s Oct. 26 progress letter sets out reductions in central government running costs that rise to €3.49 billion a year from 2035. The letter translates the savings into a rough estimate of about 25,000 fewer full-time positions. It lists AI alongside simpler rules, shared desks, standardised information technology and joint purchasing as possible tools. The government plans to publish AI targets and savings plans in the first quarter of 2027.
The letter also cautions against assuming that AI will deliver large staffing savings. It says earlier rounds of digitisation and automation did not produce the promised reductions in staff and costs, and that new tasks and contractors followed. It estimates that general efficiency measures would cover about one-third of the planned savings; the remaining two-thirds would require decisions about which activities government stops. The plan is therefore a budget and staffing programme in which AI is one proposed lever, not a stated 25,000-job cut caused by AI.
In Britain, the Home Office uses two tools built on OpenAI models through Microsoft’s Azure cloud, according to documents obtained by digital rights group Open Rights Group (ORG). One tool summarises asylum interview transcripts; another searches official country information. ORG reported that a Home Office spot check found 45 of 203 summaries failed an initial review and 30 failed a second review. Those rates—22% and 15%, respectively—were above the 9% in the pilot. The department says trained caseworkers make decisions and the tools do not decide asylum claims.
The same source material says applicants are not specifically told when AI has been used to consider their claim, and cannot see or correct the interview summary. It estimates only about 0.3% of summaries were checked in the period discussed. Separately, a UK health department review identified 90,835 of 118,756 responses to a 2023 consultation as bots and excluded them. The source says removing those responses changed the overall result, though it does not provide further detail here. Sweden’s Government Offices, meanwhile, blocked browser access to public ChatGPT and Claude on staff computers from Oct. 26, directing staff to approved tools instead.
Infographic 01 · The Netherlands
25,000 fewer civil servants. AI is one tool, not the reason.
The driver is a budget cut. The government plans to use AI, and warns against expecting too much from it.
The cut grows every year
Cuts to the running costs of central government, € billion per year.
Police and defence are exempt. Many organisations face about 20% fewer staff, some up to 45%.
Where the money must come from
The letter’s own split of the cut.
AI is one of several tools inside both parts, next to simpler rules, standard IT, shared desks and joint purchasing.
AI in the plan
- “Ambitious targets and savings plans” for AI in Q1 2027
- Purchasing to examine “systematically” where AI fits
- Conditions: security, data quality, sovereignty
Also in the plan
- At least 500 rules a year scrapped or simplified
- One desk per two employees
- Fewer advisory bodies (28 councils, ~150 committees)
The letter’s own warning
- Past digitisation did not deliver the promised staff cuts
- It brought new tasks and more contractors
- Cuts may affect service and continuity
“Do not overestimate the staffing effects of digitalisation and AI.”Lesson 2 in the Dutch government’s own letter, 26 Oct 2026 · translation: Thorsten Meyer
Sources: BZK, Voortgangsbrief Slagvaardige Overheid + annex, 26 Oct 2026; Rijksoverheid news release, 26 Oct 2026. 25,000 ÷ 160,016 ≈ 1 in 6 and bar scaling: Thorsten Meyer.
Powered by Thorsten Meyer AI
AI’s Role in Public Decisions
These cases matter because government AI can affect public-service capacity, high-stakes casework and access to decision-making. The Dutch plan links technology to a long-term effort to reduce costs and staffing, but its own letter warns that automation has not reliably delivered those savings before. That distinction matters: presenting the staffing estimate as an AI-driven job cut would overstate what the government has said.
The UK asylum example raises a different concern. Even when officials retain formal responsibility, AI-generated summaries may shape how a caseworker reads a person’s account. The reported gap between pilot and live spot-check results, combined with limited checking and no direct notice to applicants, makes accuracy and transparency material issues. Bot responses to a health consultation raise a related question: whether public submissions reflect people’s views or automated activity. Sweden’s restriction shows that governments are also making choices about which tools staff may use and under what data and contract conditions.
Products Worth Considering
FLEXISPOT EN1 Electric Standing Desk, 48 x 24 in, Black
HP Sprocket Photo Booth Machine Instant Color Photo Printer with 10.1 Touchscreen, LED Ring Light, 250+ Frames, 3×4 ZINK Paper, QR Code Shar
Dell XPS 13 9350 AI Professional Laptop 13.4 120Hz FHD+ Display
As an affiliate, we earn on qualifying purchases.
Three Different Government Approaches
The Dutch letter covers ministries, inspectorates and agencies subject to the savings programme; police and defence are exempt. Those organisations employed 160,016 full-time equivalents at the end of 2025, according to the letter. The proposed reduction is a rough estimate, and the stated savings rise over time rather than taking effect all at once.
In the UK, the Home Office tools were introduced on different schedules: the policy-search tool reached decision-makers in July 2025, and the summarisation tool reached caseworkers in April 2026. ORG published its report on Oct. 12 after parliamentary questions and freedom-of-information requests. The source material also says the tools are absent from the government’s algorithmic transparency register. Sweden’s restriction is narrower than a ban on AI: staff are to use approved alternatives, and the measure concerns the public browser versions of two products.
The source describes the EU AI Act as treating AI used in asylum decisions and access to public services as high-risk. It says that, following a digital omnibus measure, relevant duties are due to begin on Dec. 2, 2027, rather than in 2026. The material does not spell out how every duty would apply to each tool discussed.
“Do not overestimate the staffing effects of digitalisation and AI.”
— The Dutch government’s Oct. 26 progress letter
Accuracy, Oversight and Scale
Important details remain unsettled. The Dutch figure is a rough staffing estimate tied to a wider cost-cutting plan; the letter does not establish how many positions, if any, would be removed specifically because of AI. The government’s planned AI targets and savings plans are not due until the first quarter of 2027.
For the UK asylum tools, the reported spot check covered 203 summaries, and the source does not establish that the sample represents all cases or explain the full consequences of each error. The Home Office’s position is that caseworkers make decisions, but the material does not clarify how applicants could challenge an inaccurate summary or whether the tools’ performance will be independently tested. The inspector asked ORG for information while planning future inspections; the source says no probe had been confirmed.
The health consultation figures show the department’s classification and exclusion of responses, but the source does not explain the detection method or provide enough detail to assess the disputed or borderline cases. It also does not identify how the consultation’s conclusion changed. More broadly, the examples do not establish a single Europe-wide pattern: they involve different services, rules and stages of AI adoption.
Plans and Scrutiny Ahead
The Dutch government is expected to publish its AI targets and savings plans in early 2027, giving parliament and the public more detail on where it expects AI to contribute and how staffing reductions would be allocated. The cost savings are scheduled to grow through 2035, while the letter says some cuts will require choices about the scope of government work.
In Britain, the Independent Chief Inspector of Borders and Immigration has asked ORG for further information while preparing next year’s inspections, according to WIRED’s Oct. 27 reporting cited in the source material. The inspector had not confirmed a formal investigation. Questions around accuracy, disclosure, oversight and inclusion in the transparency register remain open in the material available.
Sweden’s staff restriction took effect on Oct. 26, with approved tools as the permitted alternative. The next stated EU milestone is Dec. 2, 2027, when the source says relevant AI Act duties are due to begin. Until then, the three national examples point to separate decisions rather than a settled common model for public-sector AI.
Key Questions
Is the Netherlands cutting 25,000 jobs because of AI?
No. The government’s letter gives a rough estimate of about 25,000 fewer full-time positions as part of a broader programme to reduce central government costs. It lists AI among several possible tools and warns against overestimating its staffing effects.
Does AI decide UK asylum claims?
The Home Office says trained caseworkers make every decision and that its tools do not make decisions on claims. One tool summarises interview transcripts and another searches official country information, so the reported concern is how their output may support casework.
How often did the asylum summaries fail the reported checks?
In a Home Office spot check of 203 summaries, 45 failed the first review and 30 failed a second review, equivalent to 22% and 15%. The source says the pilot failure rate was 9%; it does not establish that the spot-check sample represents every case.
Has Sweden banned AI in government?
No. The Government Offices blocked staff access to the public browser versions of ChatGPT and Claude and directed staff to approved tools. The reported measure is a restriction on particular services, not a general ban on AI.
When do the relevant EU AI Act duties begin?
The source material says duties for the high-risk uses discussed, including asylum decisions and access to public services, are due to begin on Dec. 2, 2027 after a digital omnibus measure. It does not detail how each requirement applies to the individual tools.
Source: ThorstenMeyerAI.com
