🔍 Read the full analysis: A Look At Europe’s Government AI Use: Jobs, Asylum And Bot Replies 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 report published on October 29 describes four recent European government decisions or findings involving AI: Dutch plans to reduce central-government staffing, AI tools in UK asylum casework, bot-generated responses to a health consultation and Sweden’s block on public ChatGPT and Claude. The examples raise questions about workforce plans, accuracy and transparency in decisions affecting people, and how governments verify public feedback.
A report published on October 29 describes four recent European government developments involving AI: the Netherlands’ plan for about 25,000 fewer central-government jobs by 2035, UK use of AI-generated summaries in asylum casework, bot responses to a public health consultation and Sweden’s restrictions on public ChatGPT and Claude. The cases show AI entering decisions and administrative work, while leaving open questions about accuracy, disclosure and oversight.
The Dutch figure is an estimate tied to a wider spending reduction, not an announced AI-only layoff target. A parliamentary letter sent October 26 by state secretary Eric van der Burg said central-government running-cost cuts would rise to €3.49 billion a year from 2035. It estimated the cuts could correspond to about 25,000 full-time positions. AI was listed among several possible efficiency measures, alongside simpler rules, shared desks and standardised information technology.
In Britain, the Home Office uses two tools in asylum work: 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. The report says applicants are not specifically told when AI is used and cannot see or correct the generated interview summary. The Home Office says trained caseworkers make the decisions and the tools do not decide claims.
ORG’s figures indicate that, in a 2026 review of 203 summaries, 45 failed an initial expert review and 30 failed a second review. The report compares those results with a 9% inaccuracy or incompleteness rate in the 2025 pilot. It also says roughly 203 summaries were checked against about 70,000 cases processed from April to July, or around 0.3%. Separately, the report says the UK health department flagged 90,835 of 118,756 responses to a 2023 consultation as bots and excluded them. Sweden’s Government Offices blocked staff access to public versions of ChatGPT and Claude on October 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
Together, the cases show that public-sector AI is not just a matter of adopting software. It can affect staffing assumptions, the records used in asylum decisions and the way governments interpret consultation responses. Errors or incomplete records may matter especially when a person cannot see the material or correct it, even if a human official retains formal responsibility for the decision.
The consultation figures point to a different risk: automated or bot-generated participation can distort the apparent weight of public opinion. The department’s decision to exclude flagged responses changed the reported outcome, according to the source report. The basis and method used to classify those replies matter to how readers assess that conclusion.
There is also a gap between projected savings and demonstrated results. The Dutch government itself warns against assuming digitalisation will produce large staffing reductions. Sweden’s restriction, meanwhile, is described as a contract and data-use rule, not a ban on government AI. These distinctions are relevant to workers, asylum applicants and anyone responding to public consultations.
Products Worth Considering
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
Amazon Kindle Paperwhite (12th Generation, 2024), 16 GB, 7-inch Display, Jade
As an affiliate, we earn on qualifying purchases.
Four Different Government Approaches
The Dutch estimate applies to ministries, inspectorates and agencies covered by the spending reductions. Those organisations employed 160,016 full-time equivalents at the end of 2025, according to the parliamentary letter; police and defence are exempt. The letter describes the job conversion as a rough estimate and says efficiency gains alone would cover only about one-third of the cuts. The remaining reductions require decisions about which government activities to stop.
ORG published its account of the British asylum tools on October 12 after six months of parliamentary questions and freedom-of-information requests. The tools were rolled out to decision-makers and caseworkers in 2025 and 2026. The source report says the Home Office’s own playbook calls for relevant tools to appear in an algorithmic transparency register, but the two asylum tools were not listed when ORG checked.
The developments also sit amid European regulation. The source report says the EU AI Act classifies AI used in asylum decisions and access to public services as high-risk, with relevant duties scheduled to begin on December 2, 2027 following changes described as the digital omnibus. That future date does not resolve the immediate questions about how the UK tools are monitored or how the Dutch government will translate its savings plans into staffing decisions.
Accuracy, Disclosure and Oversight
The source material does not establish how often errors in the asylum summaries altered a case outcome, or whether applicants can challenge an error through another process. It also does not provide enough detail to assess how the 203 summaries were selected for review, or whether the spot-check sample represents all cases. The reported failure rates should therefore be read as results from the cited checks, not as a definitive error rate for every summary.
The Home Office’s account that caseworkers make decisions does not answer how much staff rely on the summaries or how they handle discrepancies. The report says the Independent Chief Inspector of Borders and Immigration asked ORG for more information while planning next year’s inspections, but the inspector did not confirm a probe.
For the consultation, the source does not detail the department’s detection criteria, the number of bot responses that may have been misclassified, or the precise calculation behind the changed result. In the Netherlands, the final distribution of reductions among agencies and the specific contribution expected from AI remain pending. Sweden’s block concerns public versions of two services; it does not establish that all AI use by government staff has stopped.
Plans, Inspections and New Rules
The Dutch cabinet is expected to publish AI targets and savings plans in the first quarter of 2027. Those plans should clarify where officials expect AI to contribute and how the estimated workforce reductions will be allocated. The government’s warning about past digitalisation efforts will be a relevant measure of whether its projections are grounded in specific changes or broad efficiency assumptions.
In Britain, the next reported step is the Independent Chief Inspector’s planned inspections next year. The inspector has requested information from ORG, but the source report does not say whether the asylum tools will be formally examined or what any inspection will conclude. Further information on transparency-register entries, checking rates and applicants’ ability to review summaries could clarify how the system is governed.
The EU AI Act duties cited in the report are scheduled to apply from December 2, 2027. Until then, the four cases offer separate tests of public-sector practice: whether governments can substantiate AI-related savings, detect and correct errors in consequential administrative work, distinguish genuine public input from automated submissions and set clear rules for staff use of commercial tools.
Key Questions
Is the Netherlands cutting 25,000 jobs because of AI?
No. The rough estimate of about 25,000 positions is linked to wider central-government spending cuts. AI is one of several possible efficiency measures, and the Dutch letter cautions against overestimating its effect on staffing.
Do AI tools decide UK asylum claims?
The Home Office says trained caseworkers make decisions and that its tools do not decide claims. One tool summarises asylum interview transcripts; another searches official country information. The source report says applicants are not specifically told when AI is used.
What did the UK asylum summary checks find?
In a 2026 spot check of 203 summaries, 45 failed the first review and 30 failed a second review, according to figures reported by ORG. The source compares this with a 9% rate in the 2025 pilot; the spot-check results do not establish the error rate across all cases.
Did Sweden ban AI in government?
No. The Government Offices blocked staff access to the public versions of ChatGPT and Claude from October 26, with approved tools available instead. The source describes the move as a rule about contracts and data, not a general ban on AI.
When do the EU AI Act duties cited in the report begin?
The source report says duties for the high-risk uses it identifies, including asylum decisions and access to public services, are scheduled to begin on December 2, 2027, after the changes it calls the digital omnibus.
Source: ThorstenMeyerAI.com
