🔍 Read the full analysis: Europe’s Governments And AI: The Work, The Files And The Bot Replies on ThorstenMeyerAI.com
Get audio and creator gear 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 report published by ThorstenMeyerAI.com on October 29 describes several recent European government decisions involving AI: Dutch plans to cut central-government running costs and jobs, UK use of AI tools in asylum casework, and Sweden blocking public ChatGPT and Claude accounts on staff computers. It also reports that UK officials excluded 90,835 responses to a 2023 health consultation as bot-generated. The cases raise questions about workforce planning, oversight of AI in decisions affecting people, and how public consultation responses are verified.
A report published by ThorstenMeyerAI.com on October 29 details how governments in the Netherlands, Britain and Sweden are handling AI in public administration, from workforce planning to asylum casework and staff access to chatbots. The cases also include a UK health consultation in which officials classified 90,835 of 118,756 responses as bot-generated, illustrating how AI use and suspected automated participation can affect public services and government decisions.
In a letter sent to parliament on October 26, Dutch State Secretary Eric van der Burg set out plans to reduce central-government running costs, with annual cuts rising from €901 million in 2027 to €3.49 billion from 2035. The letter gives a rough estimate of around 25,000 fewer full-time positions. That estimate is not an AI-specific target: the government lists AI alongside other measures, including shared workspaces, standardised IT and joint purchasing. Police and defence are exempt from the cuts described.
The Dutch letter also cautions against assuming that automation will deliver large staffing savings. It says earlier digitisation and automation did not produce the expected reductions in staffing and costs, and that new tasks and contractor use followed. The government expects general efficiency gains to cover about one-third of the planned savings; the rest, the letter says, will require decisions about which activities the state stops doing. AI targets and savings plans are due in the first quarter of 2027.
In Britain, the Home Office uses two tools based on OpenAI models for asylum casework, according to documents obtained by digital rights group Open Rights Group (ORG). One summarises asylum interview transcripts; the other searches official country information. ORG says applicants are not specifically told when AI is used and cannot see or correct the summaries. The Home Office says trained caseworkers make all decisions and that the tools do not decide claims. ORG also reports that a Home Office review found errors in 15% to 22% of sampled summaries, compared with 9% in the pilot. The sample covered 203 summaries, while the source estimates about 17,515 cases pass through the summarisation tool each month.
The report describes two further developments. In a 2023 UK health consultation, the health department flagged and excluded 90,835 of 118,756 replies as bots, according to the source; the source says the consultation result changed when those responses were removed. In Sweden, the Government Offices blocked browser access to public versions of ChatGPT and Claude on staff computers from October 26, directing staff to approved tools instead. The report characterises the move as a restriction based on contracts and data handling, not a general ban on AI.
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 Services
These developments involve different uses of AI, but each can affect how the public experiences government. Dutch staffing plans could reshape public administration, though the government has not attributed the estimated job reduction to AI alone. In asylum casework, summaries may influence the records that officials consult when reviewing a person’s account, making accuracy, disclosure and routes to correction important oversight questions.
The consultation figures point to a related challenge: public agencies must distinguish genuine public views from automated submissions without distorting the response they are meant to measure. The reported change in the consultation outcome after bot-classified replies were removed underscores the potential effect of that classification. Sweden’s restriction, meanwhile, shows one government responding to the risks of staff using public AI services by limiting access to tools it has not approved.
Together, the cases show why AI policy in government is not just about deploying software. It also involves staffing decisions, data controls, transparency, quality checks and public confidence. The report does not establish that these cases share a single policy or technical cause.
Products Worth Considering
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
Bose QuietComfort Wireless Bluetooth Headphones with Active Noise Cancellation, Mic, USB‑C Charging, Deep Bass, 24‑Hour Battery – White Smok
As an affiliate, we earn on qualifying purchases.
Three Government AI Decisions
The Netherlands’ October 26 letter is a budget and public-sector reform document. Its estimate of around 25,000 fewer full-time positions is presented as a rough translation of spending cuts into jobs. The letter names AI as one possible tool, but also warns that digitisation has not reliably delivered the staffing savings expected in earlier rounds of reform. Describing the plan simply as an AI-driven job-cut programme would go beyond what the letter says.
In the UK case, ORG’s findings followed six months of parliamentary questions and freedom-of-information requests, with its report published on October 12. The source says the asylum tools were introduced across decision-making and casework teams in 2025 and 2026. It also reports that the Home Office’s AI register does not list them, although government guidance asks departments to publish information on such systems. Separately, the source says an EU AI Act provision treats AI used in asylum decisions and access to public services as high-risk; it reports that relevant duties are scheduled to begin on December 2, 2027, after changes described as a digital omnibus.
Sweden’s action differs from the UK’s deployment: it blocks staff access to specific public chatbot versions while allowing approved tools. These examples are reported together as recent decisions, not as evidence that European governments have adopted one common approach to AI.
“Do not overestimate the staffing effects of digitalisation and AI.”
— Dutch government letter to parliament, sent by State Secretary Eric van der Burg on October 26
Oversight and Error Rates Remain Open
The reported UK error rates come from a sample of 203 summaries, not a review of every asylum case handled by the tool. The source estimates that about 0.3% of summaries were checked, but the full method and representativeness of the spot check are not set out in the material provided. It is also unclear how often errors affected a caseworker’s assessment or the outcome of a claim.
The Home Office says caseworkers make the decisions, while ORG says applicants are not told about AI use and cannot inspect or correct their summaries. The source reports that the Independent Chief Inspector of Borders and Immigration asked ORG for more information while planning inspections for the following year; it does not confirm that a formal investigation has begun. The extent of the tools’ absence from the government transparency register and any planned remedy are also not established here.
For the consultation, the source does not provide the department’s full method for identifying bot responses, nor the revised result in detail. In the Netherlands, the 25,000-position figure remains a rough estimate, and the government’s future AI savings targets have not yet been published. Sweden’s block applies to public versions of two chatbots on staff computers; the report does not describe the precise technical controls or whether the rule may change.
Plans, 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 of the government’s cost-reduction programme it expects AI to support, although the October letter cautions that efficiency alone will not cover the full savings requirement.
In Britain, the next reported development is the inspector’s planned work on border and immigration services. The Independent Chief Inspector has requested information from ORG, according to the source, but has not confirmed a probe. Further details on the tools’ accuracy, checks, transparency and use in individual cases may emerge through that process or through further disclosures from the Home Office.
The source also places the UK case against the EU AI Act timetable, reporting that relevant high-risk-system duties are set to apply from December 2, 2027. How those rules will apply to specific systems and government processes is not resolved by the material provided. The immediate open questions are whether the UK tools will receive broader scrutiny, how the department handles errors and disclosure, and how agencies validate public consultation responses.
Key Questions
Is the Netherlands cutting 25,000 jobs because of AI?
No. The Dutch letter gives a rough estimate of around 25,000 fewer full-time positions as a translation of planned cost reductions. It lists AI as one of several possible tools, and warns against overestimating the staffing savings from automation.
What does the UK Home Office use AI for in asylum cases?
According to documents obtained by ORG, one tool summarises asylum interview transcripts and another searches official country information. The Home Office says trained caseworkers make the decisions and the tools do not decide claims.
How many asylum summaries were reported as inaccurate or incomplete?
The source says a Home Office pilot found 9% of summaries inaccurate or incomplete. In a later spot check of 203 summaries, 22% failed the first review and 15% failed a second review. Those figures come from a sample, not every case.
Why were UK health consultation replies excluded?
The report says the health department classified 90,835 of 118,756 replies to a 2023 consultation as bot-generated and excluded them. The source says the result changed when those replies were removed, but does not provide the department’s full detection method or detailed revised result.
Has Sweden banned AI tools in government?
No general ban is described. From October 26, Sweden’s Government Offices blocked browser versions of ChatGPT and Claude on staff computers and directed staff to approved tools. The source says the restriction concerns tool approval, contracts and data handling.
Source: ThorstenMeyerAI.com
