🔍 Read the full analysis: AI’s Expanding Footprint In Europe’s Public Jobs And Asylum Work on ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
A Dutch government plan links a projected reduction of about 25,000 central-government jobs by 2035 partly to AI, while warning that automation may not deliver large staffing savings. In the UK, a rights group says asylum interview summaries made with AI had higher error rates in a Home Office check than in a pilot; a health consultation also received a large number of responses flagged as bots. Sweden has blocked public ChatGPT and Claude on government computers, while the extent of future scrutiny and safeguards remains unsettled.
European governments are taking sharply different steps on artificial intelligence: the Netherlands has outlined a plan for about 25,000 fewer central-government full-time jobs by 2035, Britain is using AI to assist asylum caseworkers, and Sweden has blocked public versions of ChatGPT and Claude on staff computers. The cases put staffing, the handling of asylum claims and the integrity of public consultation at the centre of debates over AI in government.
The Dutch figure is an estimate attached to a wider budget reduction, not a target attributed solely to AI. In a progress letter sent to parliament on October 26, State Secretary Eric van der Burg said cuts to central-government running costs are set to rise to €3.49 billion a year from 2035. The letter describes the translation into jobs as a rough estimate. It lists AI alongside other measures, including simpler rules, shared desks, standardised information technology and joint purchasing.
The letter also cautions against assuming that digitalisation will automatically reduce headcount. It says earlier automation did not produce promised savings in staff and costs, and sometimes brought new tasks and greater use of contractors. The government expects general efficiency gains to cover about a third of the planned cuts; the rest, the letter says, requires decisions about which activities the state will stop carrying out.
In Britain, the Home Office uses two AI tools in asylum work: one summarises interview transcripts and another searches official country information. The Open Rights Group (ORG), a digital-rights organisation, says both run on OpenAI models through Microsoft’s Azure cloud. The Home Office says trained staff make all decisions and the tools do not decide claims. Applicants are not specifically told AI was used, according to a written response from Minister Anna Turley, and cannot see or correct the interview summary, ORG reports.
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 Enters High-Stakes Public Decisions
The cases matter because public agencies are applying AI to work with direct consequences for people, from employment and public spending to decisions involving asylum claims. A summary that omits or misstates part of an interview could affect how a caseworker understands an applicant’s account, even if a human official retains formal responsibility for the decision. The reported gap between pilot and live-use error rates makes the quality of review and the ability to correct errors central questions.
AI use in government also raises questions about accountability. If applicants are not told when AI has been used and cannot check a summary of their own interview, they may have limited ability to identify errors. Separately, the UK health consultation illustrates a different risk: automated submissions can affect how officials interpret public input. The department flagged a large share of responses as bots and excluded them, but the source account does not establish how every response was assessed.
For public employees, the Dutch plan shows why claims about AI-driven job cuts need careful framing. The planned reduction is tied to budget cuts and several efficiency measures, while the government itself warns that AI’s staffing effects may be limited. Sweden’s restrictions, meanwhile, show that one government response is to control which services staff can use rather than to reject AI tools altogether.
Products Worth Considering
Amazon Kindle Scribe Colorsoft 64GB (1st Generation)
Apple iPad Air 11-inch (M2), 512GB, Wi-Fi, Space Gray
Logitech MX Keys S Wireless Keyboard, Graphite
As an affiliate, we earn on qualifying purchases.
Three Different Government Responses
The Dutch letter covers ministries, inspectorates and agencies within the scope of the savings plan; police and defence are exempt. Those organisations employed 160,016 full-time equivalents at the end of 2025, according to the letter. It anticipates reductions of about 20% in many organisations and up to 45% in some. The cabinet says it will publish AI targets and savings plans in the first quarter of 2027.
ORG published its account of the UK asylum tools on October 12, after six months of parliamentary questions and freedom-of-information requests. It says the summarisation tool was rolled out to all caseworkers in April 2026 and the policy-search tool reached all decision-makers in July 2025. In a Home Office pilot, 9% of summaries were inaccurate or incomplete. In a later spot check of 203 summaries, Home Office experts rejected 45 at a first review and 30 at a second, corresponding to 22% and 15% of the sample. ORG estimates that the 203 checked summaries represented about 0.3% of roughly 70,000 cases processed from April to July.
A separate UK health consultation received 118,756 replies in 2023. The health department flagged 90,835—76%—as bot responses and excluded them, according to the source account. It says the consultation result changed when those responses were removed. In Sweden, the Government Offices blocked browser versions of ChatGPT and Claude on staff computers from October 26. Staff are directed to approved tools; the measure is a restriction on public services and data handling, not a general ban on AI.
Questions Over Error Checks and Jobs
It is not clear whether the Home Office will change how applicants are informed or allow them to review AI-generated summaries. ORG says the tools are not listed in the government’s algorithmic transparency register, while the Home Office maintains that trained staff make the decisions. The source report does not establish whether the department has since responded to the group’s findings or expanded its checking beyond the cited sample.
On October 27, WIRED reported that the Independent Chief Inspector of Borders and Immigration had asked ORG for more information while planning next year’s inspections. The inspector had not confirmed a probe, so the scope and timing of any examination remain uncertain. The reported error rates also come from different exercises—a 2025 pilot and a 2026 spot check—rather than a continuous, independently verified audit.
The Dutch figure is a rough job estimate within a broader savings programme, and the government has not yet published its detailed AI targets. The health consultation account gives the number of responses flagged as bots but does not provide the full criteria or breakdown of how those classifications were made. The source material also does not specify whether Sweden’s restriction will be reviewed or expanded.
Inspections and Savings Plans Ahead
The next stated milestone in the Netherlands is the cabinet’s planned publication of AI targets and savings plans in the first quarter of 2027. Those plans should clarify what work the government expects AI to support and how that relates to the projected staffing reductions. The progress letter says efficiency alone will not cover the full budget cut, leaving decisions about the scope of government work unresolved.
In Britain, the inspector’s planned work next year may bring further scrutiny of the asylum tools, although no formal investigation was confirmed in the reported update. The key questions will be whether checking becomes more frequent, whether applicants receive disclosure or correction options, and how officials document AI’s contribution while retaining responsibility for decisions.
The source report also says the EU AI Act treats systems used in asylum decisions and access to public services as high-risk. It gives December 2, 2027 as the start date for those duties after a digital omnibus measure. How those requirements will apply to the specific tools described, and what practical changes agencies will make before then, remains to be seen.
Key Questions
Is the Netherlands cutting 25,000 jobs because of AI?
No. The government letter presents about 25,000 fewer full-time positions as a rough estimate connected to a wider budget-cutting plan. AI is one of several measures, and the letter warns against overestimating its effect on staffing.
Does AI decide UK asylum claims?
The Home Office says the tools do not make decisions on claims and that trained caseworkers decide each case. One tool summarises interview transcripts and another searches official country information, according to the source report.
How many AI asylum summaries were checked?
ORG says Home Office experts checked 203 summaries in a spot check. It estimates that this was about 0.3% of roughly 70,000 cases processed from April to July; the estimate and its denominator are reported by ORG.
Has the UK borders inspector opened an investigation?
That has not been confirmed. WIRED reported that the inspector asked ORG for information while planning next year’s inspections, but the inspector had not confirmed a probe.
Has Sweden banned AI in government?
No. The reported measure blocks browser versions of ChatGPT and Claude on Government Offices computers and directs staff to approved tools. It is a restriction on those public services, not a general ban on AI.
Source: ThorstenMeyerAI.com
