🔍 Read the full analysis: AI’s Paperwork Wave Is Challenging British Bureaucracy on ThorstenMeyerAI.com
TL;DR
A Sky News report published October 20 says written complaints, freedom-of-information requests and tribunal claims have risen across several UK public services since ChatGPT launched. The figures show a broad pattern, but do not prove AI caused the increases; officials and experts also warn that many submissions may reflect valid claims made easier to pursue.
A Sky News report published October 20 found that complaints, freedom-of-information requests and employment tribunal claims have risen across several UK public services since ChatGPT launched in November 2022. The pattern is putting pressure on bodies that handle public business through written forms and correspondence, although the figures do not establish that AI caused the increases.
Using data seen by Sky News technology correspondent Rowland Manthorpe, the report compared changes across nine categories. Household complaints about water companies rose 144% since ChatGPT’s launch, while complaints about English local authorities rose 78% and written complaints to GP practices rose 71%. The report also recorded increases in complaints to the Department for Work and Pensions (64%), individual employment tribunal claims (62%), legal-services complaints (49%), written NHS complaints (35%), police complaints (34%) and complaints about public services (23%).
These are reported changes since ChatGPT launched, not proof that AI produced each submission or caused each rise. The report says the channels have different possible drivers; water companies, for example, have faced difficult years. Its central observation is that several unrelated channels were relatively flat for years before rising from late 2022 and accelerating afterward. The comparison does not, on its own, separate AI’s role from other causes.
Manthorpe also tested 100 recent FOI requests with an AI-text checker. It labelled 89% fully AI-written and 9% human; a smaller 2019 sample was labelled entirely human. Sky News cautioned that AI detectors are unreliable and that this limited test cannot establish either authorship or causation. Separately, an academic survey cited in the report found 58% of surveyed civil servants said AI-generated content from the public had a significant impact on their work.
Infographic 01 · The paperwork curve
Flat for a decade. Then ChatGPT made paperwork cheap.
Across almost every channel where people write to the British state, volume has risen since November 2022. Much of it is justified.
Growth since ChatGPT launched
Change per channel since November 2022, as measured by Sky News. Bar length on a 0–150% scale.
Each channel has its own causes. The repeated shape, flat then accelerating from late 2022, is the signal.
Not just Britain
Other measurements of the same trend.
- UK Housing Ombudsman: about 2,600 complaints in 2022, just over 7,000 last year (TechCrunch).
- US Consumer Financial Protection Bureau: complaints up fivefold over the same period.
- 84 cases in 11 jurisdictions of possible “agentic flooding”, most flat before 2022 (Chris Schmitz).
- ICO guidance, May 2026: AI-written FOI requests are valid; authorities may ask for clarification.
Schmitz: “The vast majority of cases we find are people who are entitled to claim for something, claiming for that thing.”
Sources: Sky News, Why AI could break Britain’s bureaucracy (20 Oct 2026); TechCrunch (10 Sep 2026); ICO (May 2026).
Powered by Thorsten Meyer AI
Pressure on Paper-Based Services
The reported increases matter because many public services still depend on people explaining a problem in writing, then staff reading and assessing each submission. If AI makes drafting a complaint or appeal faster and easier, more people may be able to use processes that previously demanded time, confidence or writing skills. The same change can also add to workloads and lengthen waits when public bodies lack staff or systems to process more correspondence.
The key policy choice is not simply whether to accept AI-written text. The report describes evidence that a substantial share of additional claims may be legitimate. At the DWP, the proportion of complaints upheld reportedly fell from the 50s to close to 40%, but because total complaints rose, the number upheld may still be higher. The underlying counts are not supplied here, so that absolute comparison cannot be checked from the reported percentages alone. Treating volume as proof of abuse could restrict access for people with valid claims.
Tom Loosemore, founder of the Government Digital Service, argues the growth could expose a weakness in services built around essays and forms. He says the current model for benefits may not survive without fundamental changes. That is his assessment, not a confirmed prediction, but it points to the practical stakes: public agencies may need ways to process more requests without making access harder.
As an affiliate, we earn on qualifying purchases.
From ChatGPT to Automated Claims
The report traces the rise to the period after ChatGPT launched in November 2022, when members of the public could use chatbots to draft formal letters and requests. It says most submissions still involve people copying text into and out of a chatbot. The technology can lower the effort required to produce a polished complaint, but the reported data cannot show how much of the overall growth comes from AI rather than service problems, policy changes or other factors.
Similar patterns have been reported elsewhere. A paper by researcher Chris Schmitz, covered by TechCrunch in September, identified 84 possible cases of what he calls “agentic flooding” across 11 jurisdictions. The source material says most were roughly flat before 2022 and rose afterward. In the UK, the Housing Ombudsman handled about 2,600 complaints in 2022 and just over 7,000 last year; the US Consumer Financial Protection Bureau’s complaints grew fivefold, according to the material.
In May, the Information Commissioner’s Office issued guidance on AI-generated FOI requests after public authorities reported higher volumes and more complex submissions, including requests that misquoted legislation or needed clarification. The ICO said an AI-written request remains valid, authorities do not need to determine who or what wrote it, and they can seek clarification when a request is unclear.
“It’s like being bombarded with correspondence at speed.”
— Civil servants quoted in the academic survey cited by Sky News
As an affiliate, we earn on qualifying purchases.
What the Figures Cannot Establish
The reported timing and cross-channel pattern do not prove that AI caused the increases. Each category may have separate drivers, and the source material does not provide the full underlying datasets, exact measurement windows for every comparison, or a method for isolating AI’s effect. The AI-detector test is also too limited and unreliable to confirm authorship across FOI requests.
It remains unclear how much of the additional correspondence is AI-assisted, how much is generated autonomously, and how much reflects previously unmet demand. The survey finding that 58% of civil servants report a significant impact describes respondents’ experiences; it does not quantify the total volume of AI-written submissions or establish how that workload affects response times and outcomes.
As an affiliate, we earn on qualifying purchases.
A Shift From Essays to Data
The next pressure point may be AI agents that can submit requests or claims on a person’s behalf, rather than only drafting text for them. Manthorpe’s report demonstrates a Meta agent service and raises examples such as appealing parking tickets or claiming rail-delay compensation automatically. These are prospective examples, not evidence that such automated activity is already widespread.
Loosemore’s proposed response is to build services around structured data rather than essays, so agencies can collect information in a more consistent format. The source material does not identify a government plan or timetable for such a redesign. Public bodies will continue to apply existing rules, including the ICO’s guidance on valid FOI requests, while the scale and effects of AI-assisted submissions develop.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did the Sky News report find?
It reported increases since ChatGPT launched across several written channels, including complaints to public bodies, FOI requests and employment tribunal claims. The figures show a pattern but do not prove that AI caused it.
Does the report show that most new complaints are written by AI?
No. A small test of 100 recent FOI requests found that an AI detector labelled 89% fully AI-written, but the report cautioned that detectors are unreliable and the sample cannot establish authorship across all complaints.
Are AI-generated FOI requests valid?
The Information Commissioner’s Office said AI-written requests remain valid. Authorities do not need to identify whether AI was used, though they can ask for clarification if a request is unclear.
Do higher complaint numbers mean more abuse of the system?
Not necessarily. The report cites research suggesting many claims are legitimate, while also noting that complaint outcomes vary. The available figures do not establish the share of increased submissions that are valid.
What changes could public services make?
Tom Loosemore argues for services that collect structured information rather than relying on long written explanations. The source material does not identify an adopted government policy or implementation date.
Source: ThorstenMeyerAI.com
