🔍 Read the full analysis: AI Enters European Government: What The Job And Asylum Figures Show on ThorstenMeyerAI.com
Get the little things that make your day 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 in the Netherlands, Britain and Sweden show different approaches to AI in government: planned workforce reductions, AI-assisted asylum casework and restrictions on public chatbot tools. A UK health consultation also received a large volume of responses identified by the department as bots. The scale of AI’s role in Dutch job cuts, the effects of errors in asylum summaries and the origins of flagged consultation replies remain important questions.
Recent developments in the Netherlands, Britain and Sweden show governments using, reviewing or restricting artificial intelligence in public administration, from Dutch plans for roughly 25,000 fewer central-government jobs by 2035 to UK AI-generated asylum interview summaries that failed up to 22% of a sample review. The cases matter because they put staffing, individual rights and public participation at stake, while leaving open how much AI is responsible for the changes.
The Dutch government’s October 26 progress letter sets out spending reductions that rise to €3.49 billion a year from 2035. It translates those cuts into a rough estimate of about 25,000 full-time positions, around one-sixth of the government workforce covered by the plan. The letter lists AI among several possible efficiency measures, alongside simpler rules, shared workspaces, standardised IT and joint purchasing. Police and defence are exempt from the cuts.
The letter also cautions against treating AI as the main source of savings. It says earlier digitalisation and automation did not produce promised reductions in staff and costs, and warns against overestimating AI’s workforce effects. The government expects general efficiency measures to cover about one-third of the planned savings; the rest would require decisions about which government activities to stop. It plans to publish AI targets and savings plans in the first quarter of 2027.
In Britain, the Home Office uses two tools in asylum casework: one summarises interview transcripts and another searches official country information. Documents obtained by the Open Rights Group (ORG) indicate both use OpenAI models through Microsoft’s Azure cloud. A Home Office spot check found 45 of 203 summaries failed an initial review and 30 failed a second review. The department’s 2025 pilot had recorded a 9% rate of summaries judged inaccurate or incomplete. The Home Office says trained staff make all decisions and the tools do not decide claims.
ORG says applicants are not specifically told when AI is used and cannot see or correct the summary of their interview. It estimates roughly 70,000 summaries were produced from April through July, while experts checked 203—about 0.3%. The group also says the tools are absent from the government’s algorithmic transparency register. Separately, Britain’s health department flagged 90,835 of 118,756 responses to a 2023 consultation as bots and excluded them. The source report says removing those responses changed the consultation’s result, but the available material does not establish who or what generated each flagged reply.
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 examples concern more than whether AI can reduce administrative workload. In asylum casework, a summary may shape how a decision-maker reads a person’s account of why they fled, even if the software does not formally make the decision. A summary error could matter if important details are omitted or misstated. The reported review figures raise questions about accuracy and oversight, but do not show how often an error affected an asylum outcome.
The Dutch plan also shows why workforce announcements need careful wording. The estimated job reduction follows a broad budget-cutting programme, and the government describes AI as one possible tool rather than the sole cause. In the consultation case, excluding bot responses affected the recorded public input. That makes transparent methods and clear reporting important for understanding whose views were counted. Sweden’s restrictions, meanwhile, show that governments may limit public AI services on staff devices without rejecting all workplace AI.
Products Worth Considering
TheraGun Relief Massage Gun – 3 Speed, Black
Oral-B iO Series 3 Rechargeable Electric Toothbrush for Adults, Black
Philips Sonicare DiamondClean Smart 9300 Electric Toothbrush, Black
As an affiliate, we earn on qualifying purchases.
Three Governments, Different Controls
The Dutch figures come from a progress letter on the cabinet’s “decisive government” programme. The agencies and ministries covered by the cuts employed 160,016 full-time equivalents on December 31, 2025, according to the letter. The 25,000 figure is explicitly described as a rough estimate, not a final staffing plan. The letter expects reductions to vary among organisations, with some facing cuts of up to 45%.
In the UK, the asylum summary tool reached all caseworkers in April 2026, while the policy-search tool reached all decision-makers in July 2025, according to the source material. ORG published its findings on October 12 after making parliamentary questions and freedom-of-information requests. 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 formal investigation.
Sweden’s Government Offices blocked browser versions of ChatGPT and Claude on staff computers from October 26, according to the report. Staff are directed to approved tools. The restriction concerns approved contracts and data handling, not a general ban on government use of AI. The report also says the EU AI Act classifies AI used in asylum decisions and access to public services as high-risk, with the relevant duties scheduled to begin on December 2, 2027, following changes described as the “digital omnibus.”
“Do not overestimate the staffing effects of digitalisation and AI.”
— The Dutch government’s progress letter
Accuracy, Accountability and Scale
The reported spot check does not establish the overall error rate across every asylum summary: it covered 203 summaries, and the source does not explain whether the sample represents all cases. The figures also do not show how many errors changed, delayed or otherwise affected decisions. It remains unclear whether applicants will be given a way to inspect or challenge AI-generated summaries, or whether the Home Office will list the tools in its transparency register.
The Dutch estimate is not a confirmed count of jobs that will be removed because of AI. The government has not yet published its detailed AI savings plans, and the source says two-thirds of the cuts may require decisions about which services or activities to discontinue. For the UK consultation, the material gives the number of replies flagged as bots but does not identify their sources or provide the criteria used to classify them. It also does not give enough detail to independently assess how the result changed.
Sweden’s restriction is described as applying to public browser versions of ChatGPT and Claude on Government Offices computers. The available information does not set out how long it will remain in place or which approved tools staff may use. The source likewise reports the EU AI Act timetable but does not detail how every government system will be assessed under the rules.
Reviews and Savings Plans Ahead
The Dutch government is due to publish its AI targets and savings plans in the first quarter of 2027. Those plans should clarify which public functions are expected to use AI, what savings are attributed to it and how the wider workforce reduction will be allocated. Until then, the 25,000 figure remains a rough estimate attached to the budget programme.
In Britain, the inspectorate’s next-year inspection planning may bring further scrutiny of asylum casework, but the inspector had not confirmed a probe as of the reported update. The Home Office could also face pressure to explain the sampling process, accuracy checks, applicant notification and transparency-register status. The source does not report a scheduled decision or a government commitment to change the tools.
The EU AI Act duties cited in the report are scheduled to begin on December 2, 2027. The timing gives governments a further milestone for explaining how high-risk systems are monitored, although the material does not specify what action each administration will take before then. Sweden’s approved-tool policy and the UK consultation findings may also prompt closer attention to procurement, data safeguards and methods for validating public input.
Key Questions
Is the Netherlands cutting 25,000 jobs because of AI?
No. The Dutch government’s letter estimates roughly 25,000 fewer full-time positions as a result of wider spending reductions by 2035. It lists AI as one of several possible efficiency tools and warns against overestimating AI’s staffing effects.
What did the UK review find about asylum summaries?
In a Home Office spot check of 203 summaries, 45 failed the first review and 30 failed a second review. The sample does not establish the error rate across all cases, and the reported figures do not show how many errors affected asylum outcomes.
Do AI tools decide UK asylum claims?
The Home Office says trained caseworkers make every decision and that its tools do not decide claims. One tool summarises interview transcripts; another searches official country information. The source says applicants are not specifically told when AI has been used.
Why were most responses to the UK health consultation excluded?
The health department flagged 90,835 of 118,756 replies to a 2023 consultation as bots and excluded them. The available material does not give the classification criteria or identify the sources of the flagged responses.
Has Sweden banned AI in government?
No. The reported measure blocks browser versions of ChatGPT and Claude on Government Offices staff computers from October 26. Staff are directed to approved tools; the restriction is described as a contract and data policy, not a general ban on AI.
Source: ThorstenMeyerAI.com
