🔍 Read the full analysis: Europe’s Public Sector Turns To AI For Jobs And Asylum Work on ThorstenMeyerAI.com
Get the latest gadgets 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
A Dutch government plan links a projected reduction of about 25,000 central-government jobs by 2035 to a wider spending cut, with AI named as one possible tool. In the UK, a watchdog is considering information about AI-generated asylum interview summaries after a report found more errors in a later spot check than in the pilot; Sweden has blocked public ChatGPT and Claude access on government computers.
Governments in the Netherlands, Britain and Sweden are taking different approaches to AI in public administration: Dutch officials have tied AI to a plan for fewer civil-service jobs, the UK Home Office uses AI tools in asylum casework, and Sweden has blocked public ChatGPT and Claude on staff computers. The developments raise questions about staffing, accuracy and oversight, while the available evidence does not show that AI alone is driving Dutch job reductions or making UK asylum decisions.
A Dutch government progress letter sent to parliament on 26 October sets out running-cost reductions that rise from €901 million in 2027 to €3.49 billion a year from 2035. It translates the plan into a rough estimate of about 25,000 fewer full-time positions across the central-government bodies covered. Police and defence are exempt. The letter lists AI alongside other measures, including simpler rules, shared desks, standardised information technology and joint purchasing.
The letter cautions against assuming that AI will deliver large staffing savings. It says earlier digitisation and automation did not produce the promised reductions in costs and staff, and sometimes brought new work and greater use of contractors. It estimates general efficiency measures will cover about one-third of the planned cuts; the rest would require decisions about which government activities to stop. The government plans to set AI targets and savings plans in the first quarter of 2027.
In Britain, the Home Office uses two tools built on OpenAI models through Microsoft’s Azure cloud, according to documents obtained by the digital rights group Open Rights Group (ORG). One summarises asylum interview transcripts; the other searches official country information. ORG reported that a Home Office review found errors or omissions in 15% to 22% of a sample of 203 summaries, compared with 9% during the pilot. The department says trained caseworkers make all asylum decisions and the tools do not decide claims. It has said applicants are not specifically told when AI is used, and they cannot see or correct the summary, according to the source material.
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 put different public-sector risks in view. In the Netherlands, the scale of the proposed workforce reduction is tied to a budget plan, but the government’s own warning makes it inaccurate to describe all the jobs as AI-driven cuts. The distinction matters to civil servants and the public assessing whether automation can plausibly deliver the stated savings.
In asylum casework, summaries of interviews can affect how decision-makers understand an applicant’s account, even if officials retain formal responsibility for the decision. The reported gap between pilot and later sample error rates, combined with limited checking, makes accuracy and the ability to challenge a summary central concerns. The Home Office has reported time savings from the tools, but those figures do not by themselves establish that the system improves decisions.
Sweden’s restriction illustrates a separate issue: government departments must decide which AI services staff may use and under what data and procurement rules. Meanwhile, a UK consultation case described in the source material raises concerns about automated or bot submissions affecting public feedback. Taken together, the developments concern how governments govern AI use, not just whether they adopt it.
Products Worth Considering
Bose QuietComfort Wireless Bluetooth Headphones with Active Noise Cancellation, Mic, USB‑C Charging, Deep Bass, 24‑Hour Battery – White Smok
Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Black
Amazon Kindle Paperwhite Signature Edition (12th Generation, 2024), 32 GB, Metallic Black
As an affiliate, we earn on qualifying purchases.
Three Different Government Approaches
The Dutch letter covers central-government organisations that employed 160,016 full-time equivalents at the end of 2025, according to the source material. The projected 25,000 reduction is an estimate attached to a broader programme of spending cuts, not a published count of positions already eliminated. The government has also said that efficiency gains alone will not account for all of the planned reduction.
ORG’s report on the UK tools was published on 12 October after six months of parliamentary questions and freedom-of-information requests. The asylum case summarisation tool reached all caseworkers in April 2026, while the policy-search tool reached all decision-makers in July 2025. The source says around 17,515 cases a month pass through the summary tool, while 203 summaries were checked in the later review—about 0.3% of the estimated volume from April through July.
Sweden’s Government Offices blocked browser access to public versions of ChatGPT and Claude from 26 October, according to the source. The restriction is described as a rule about contracts and data, not a general ban on AI; staff are directed to approved alternatives. The source also says EU rules classify AI used in asylum decisions and access to public services as high-risk, with relevant duties scheduled to begin on 2 December 2027 after a legislative timetable change.
“Do not overestimate the staffing effects of digitalisation and AI.”
— Dutch government progress letter to parliament
Accuracy Checks and Oversight
The reported UK spot check is a sample, and the source does not provide enough detail to establish how representative it is of all summaries or how the errors affected individual asylum decisions. It is also unclear whether the Home Office will expand its checks, notify applicants more directly or add the tools to the government’s algorithmic transparency register. ORG says neither tool appears in that register.
The Independent Chief Inspector of Borders and Immigration asked ORG for more information while planning next year’s inspections, according to a 27 October report by WIRED. The inspector has not confirmed an investigation. The source also refers to a tribunal judge’s July comments but cuts off before giving the substance, so no conclusion about that ruling can be drawn here.
In the Netherlands, the 25,000 figure remains a rough estimate, and the government has not yet published the promised AI targets and savings plans. It is not yet clear which roles or services would be affected. The source’s account of the UK health consultation says bots accounted for 90,835 of 118,756 replies, but does not provide enough detail here to independently assess the classification method or how the consultation’s revised result was used.
Plans and Inspections Ahead
The Dutch government is due to publish its AI targets and savings plans in the first quarter of 2027. Those documents should clarify which AI applications officials expect to use and how they connect to the wider spending programme. The job estimate may also change as departments work out which activities will be reduced or ended.
In Britain, the inspector’s planned work for next year could bring further scrutiny of asylum casework, but the inspector has not confirmed a probe. Any later findings would help clarify how often summaries contain errors, how staff review them and whether applicants receive ways to correct inaccuracies. The Home Office’s position remains that trained caseworkers make the decisions.
Sweden’s restrictions apply to public versions of ChatGPT and Claude on Government Offices computers; the source says approved tools remain available. The broader EU timetable cited in the source places high-risk AI obligations on 2 December 2027. Until then, the immediate developments to watch are national rules, published performance information and evidence of how AI affects actual public services.
Key Questions
Is the Netherlands cutting 25,000 jobs because of AI?
No. The government letter gives a rough estimate of about 25,000 fewer full-time positions as part of wider spending cuts. It lists AI as one tool among several and warns against overestimating AI’s 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; another searches official country information.
What did the UK review find about asylum summaries?
In a 2026 spot check of 203 summaries, Home Office experts marked 15% to 22% as inaccurate or incomplete across two reviews. The source reports a 9% error rate in the pilot, but the later check was a limited sample.
Has Britain’s border inspector opened an investigation?
Not according to the information provided. The Independent Chief Inspector of Borders and Immigration asked ORG for more information while planning next year’s inspections, but has not confirmed a probe.
Has Sweden banned AI in government?
No. The reported measure blocks public ChatGPT and Claude through browsers on Government Offices computers. It is described as a restriction tied to contracts and data, with approved tools available to staff.
Source: ThorstenMeyerAI.com
