🔍 Read the full analysis: How Europe’s Governments Are Bringing AI Into 25,000 Jobs And Asylum Files on ThorstenMeyerAI.com
Get movie nights 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 projects about 25,000 fewer central-government full-time positions by 2035, with AI among several proposed efficiency measures. In Britain, a rights group says Home Office asylum summaries showed more errors in a later check than in a pilot; Sweden has blocked public ChatGPT and Claude access on government computers. The source also reports that bots accounted for 76% of replies to a 2023 UK health consultation, according to the department.
European governments are taking sharply different approaches to artificial intelligence in public services: the Netherlands has projected about 25,000 fewer central-government full-time jobs by 2035 while listing AI as one possible efficiency tool, Britain is using AI to summarise asylum interviews, and Sweden has blocked public versions of ChatGPT and Claude on government computers. The cases raise questions about staffing, the use of automated tools in decisions affecting people, and oversight of government systems.
A Dutch government progress letter sent to parliament on October 26 links a planned reduction in central-government running costs to an estimated 25,000 full-time positions by 2035. The letter calls the jobs figure a rough estimate. Annual savings are projected to rise from €901 million in 2027 to €3.49 billion from 2035. Police and defence are exempt from the cuts described.
The Dutch letter lists AI alongside other measures, including simpler rules, shared desks, standardised information technology and joint purchasing. The cabinet says it will publish AI targets and savings plans in the first quarter of 2027. At the same time, the letter warns officials not to overestimate the employment effects of digitalisation and AI, citing past cases in which automation did not deliver promised staff or cost savings.
Britain’s Home Office uses two generative AI tools in asylum casework, according to documents obtained by the digital rights group Open Rights Group (ORG). One condenses interview transcripts; the other searches official country information. ORG says a Home Office check found 45 of 203 summaries failed an initial review and 30 failed a second review. The source reports a 9% inaccuracy or incompleteness rate in the earlier pilot. The department says trained caseworkers make decisions and that the tools do not decide claims.
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 Touches Jobs and Asylum Decisions
The Dutch estimate makes clear that public-sector AI is being discussed within a broader budget and staffing programme, not as the sole cause of planned job reductions. The government itself cautions that AI may not produce the savings sometimes expected. Its letter says general efficiency may cover only about a third of the planned reductions, with other cuts requiring decisions about which government activities to stop.
The British case has a more direct potential effect on individuals. An interview summary could shape how a caseworker reviews an asylum claim, even though the Home Office says a human makes the decision. ORG says applicants are not specifically told when AI has been used and cannot see or correct the summaries. That combination makes accuracy, transparency and the opportunity to challenge errors relevant to due process.
These examples also show that government AI policy is not simply a choice between adoption and prohibition. Sweden’s restriction applies to public chatbot versions on staff computers, while approved tools remain available. The reported UK consultation case points to a separate risk: automated or fabricated submissions can affect the apparent weight of public feedback.
Popular right now
KitchenAid Artisan 5-Quart Tilt-Head Stand Mixer, Almond Cream
Samsung 48-Inch S90H OLED 4K Smart TV (2026)
Vitamix Propel Series 510 Professional-Grade Blender, 3 Programs, Black
As an affiliate, we earn on qualifying purchases.
Three Different Government Approaches
The source material describes three distinct developments across the same week. In the Netherlands, the October 26 parliamentary letter sets out a long-term cost-cutting plan, with AI included among several proposed tools. The government’s estimate is based on a central-government workforce of 160,016 full-time equivalents at the end of 2025, excluding police and defence from the cuts. The letter says reductions at individual organisations could vary widely.
In Britain, the Home Office’s Asylum Policy Search tool was made available to decision-makers in July 2025, and its Asylum Case Summarisation tool to caseworkers in April 2026, according to the source. ORG published its findings on October 12 after seeking information through parliamentary questions and freedom-of-information requests. The source says the tools use OpenAI models through Microsoft’s Azure cloud.
The source also reports that Sweden’s Government Offices blocked browser access to public ChatGPT and Claude on October 26, directing staff to approved tools instead. This was described as a restriction tied to contracts and data handling, rather than a general ban on AI. Separately, Britain’s health department received 118,756 replies to a 2023 consultation and flagged 90,835—76%—as bots, according to the source. The department excluded those replies; the source says the result would have been different without them.
“Do not overestimate the staffing effects of digitalisation and AI.”
— The Dutch government progress letter to parliament
Accuracy Checks and Oversight Gaps
The reported Home Office accuracy figures come from a small spot check of 203 summaries, compared with an earlier pilot. The source does not provide enough detail to establish whether the checks used the same method or sampled cases in comparable ways. It also reports that about 17,515 cases a month pass through the summarisation tool, while only 203 summaries were checked in the cited review; the broader error rate is not established by that sample.
It remains unclear whether the UK’s Independent Chief Inspector of Borders and Immigration will investigate the tools. The source says the inspector asked ORG for more information while planning next year’s inspections, but did not confirm a probe. The Home Office tools are also absent from the public algorithmic transparency register, according to ORG.
The Dutch figure is a rough estimate, not a confirmed count of jobs that will be removed by AI. The government has not yet published the detailed AI targets and savings plans. The source also says EU rules classify AI used in asylum decisions and access to public services as high-risk, with relevant duties due to begin on December 2, 2027 after a legislative change; the precise obligations for each system are not detailed here.
Targets, Inspections and AI Rules
The Netherlands is expected to publish its AI targets and savings plans in the first quarter of 2027. Those plans should clarify how much the government expects AI to contribute, and how the projected reductions will be divided among organisations and other spending measures.
In Britain, the border inspector’s next inspection plans may determine whether the asylum tools receive further scrutiny. The source says the inspector requested information from ORG but had not confirmed an investigation. More detail on the Home Office’s quality controls, transparency arrangements and handling of inaccurate summaries would help establish how the tools operate in practice.
Across the EU, the source places the next major regulatory milestone on December 2, 2027, when high-risk AI duties are due to start under the revised timetable. Until then, the immediate questions are how public bodies test systems, disclose their use and allow people affected by them to identify or challenge errors.
Key Questions
Is the Netherlands cutting 25,000 jobs because of AI?
No. The Dutch letter estimates about 25,000 fewer full-time positions as part of a wider budget-cutting plan. It lists AI as one of several possible tools 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 decide claims. One tool summarises interviews and another searches official country information.
What accuracy concerns were reported about the UK summaries?
ORG reported that 45 of 203 summaries failed an initial Home Office review and 30 failed a second review. The source also cites a 9% rate in an earlier pilot; the spot check does not establish an error rate across all cases.
Has Sweden banned AI in government?
No general ban is described. The source says the Government Offices blocked browser access to public ChatGPT and Claude on staff computers from October 26, while staff use approved tools.
When are the EU high-risk AI duties due to begin?
The source gives December 2, 2027 as the start date for duties applying to high-risk AI uses such as asylum decisions and access to public services. It does not detail how each government system will be assessed.
Source: ThorstenMeyerAI.com
