How Britain’s Bureaucracy Is Handling AI’s Paperwork Flood

🔍 Read the full analysis: How Britain’s Bureaucracy Is Handling AI’s Paperwork Flood on ThorstenMeyerAI.com

A Sky News report published October 20 finds that complaints, Freedom of Information requests and employment tribunal claims have risen across multiple British public services since ChatGPT launched in November 2022. The figures show a broad increase in written submissions, while the report cautions that the trend does not establish that AI caused it; the development matters because public bodies depend on staff time to process those submissions, many of which may represent valid claims or unmet needs.

Using data seen by Sky News technology correspondent Rowland Manthorpe, the report compares reported increases across several channels since ChatGPT’s launch. Household complaints about water companies rose 144%, complaints about English local authorities rose 78%, and written complaints to GP practices increased 71%. Complaints to the Department for Work and Pensions rose 64%, while individual employment tribunal claims were up 62%. The report gives the period as since launch but does not specify a common end date or a separate baseline for every measure.

Other reported increases include 49% for complaints about legal services, 35% for written NHS complaints, 34% for police complaints and 23% for complaints about public services. The report says each area has its own possible causes, including difficulties affecting water companies. Its central observation is that multiple series, after remaining comparatively flat for years, began rising around late 2022. That timing is an association, not proof of a cause.

Manthorpe also tested 100 recent FOI requests with an AI-text detector. It marked 89% as fully AI-written and 9% as human; a smaller 2019 sample was classified as entirely human. The report stresses that AI detectors are unreliable and that this small test cannot show that AI produced the wider rise. Separately, an academic survey cited by Sky News found that 58% of civil servants surveyed said AI-generated material from the public significantly affected their work.

More Submissions, Greater Processing Pressure

The figures point to a practical strain for services that handle large volumes of letters, applications and appeals. If more people can use chatbots to draft formal submissions, staff may face more material to read and answer, including requests that are lengthy, confusing or based on misquoted rules. Civil servants quoted in the report described feeling bombarded and said the work could be exhausting and anxiety-inducing.

But reducing volume is not the same as resolving the underlying demand. The report says the proportion of DWP complaints upheld has fallen from the 50% range to close to 40%. That is a lower success share, not evidence that complaints are mostly baseless; with a much larger total, the number upheld may still be higher. Researcher Chris Schmitz, whose work the report cites, argues that most cases he examined involve people claiming something to which they are entitled. The policy choice is therefore consequential: adding barriers could also make it harder for eligible people to seek redress.

The report presents structured digital services as an alternative to processing free-form essays. If public systems collect information in consistent formats and make decisions through clear workflows, they may be better able to handle higher demand without simply discouraging people from applying or complaining. Whether that approach can be delivered across varied services is not settled.

How Officials Are Responding

In May, the Information Commissioner’s Office published guidance on AI-generated FOI requests after public authorities reported higher volumes and more complicated submissions. Some requests, the ICO said, misquoted legislation or needed substantial clarification. Its guidance states that an AI-written request remains a valid request. Authorities do not need to determine whether a person or an AI wrote it, and they can seek clarification when a request is unclear or badly formed.

The pressure is not limited to Britain. A paper by Chris Schmitz, reported by TechCrunch in September, identified 84 possible cases of “agentic flooding” across 11 jurisdictions. The source describes many as roughly flat before 2022 and rising afterward. It also cites the UK Housing Ombudsman, where complaints increased from about 2,600 in 2022 to just over 7,000 the following year, and the US Consumer Financial Protection Bureau, where complaints grew fivefold. These comparisons suggest a wider pattern, but do not establish a shared cause in every case.

Most current submissions described in the Sky News report involve people using chatbots to draft text. The next change could come from AI agents that can take actions for users, not just prepare letters. Manthorpe demonstrated a Meta agent service and raised examples such as automated parking appeals or rail delay-compensation claims. Those examples describe possible future activity, not evidence that such systems are already generating mass claims.

““It’s like being bombarded with correspondence at speed.””

— Civil servants responding to an academic survey cited by Sky News

What the Rising Counts Cannot Prove

The report shows that several types of written submissions rose over a period that began around ChatGPT’s release, but it does not demonstrate that AI caused those increases. Services have distinct circumstances, and the figures cover different channels. The source does not provide a shared measurement method, a common end date for every series or enough detail to compare the underlying populations directly.

The AI-detector test is also limited: it used a small sample, and the report warns that detector results are unreliable. It cannot reliably establish authorship across the full volume of requests. It is also unclear how many recent submissions were written by people using chatbots, produced by agents acting for them, or written without AI at all. The scale and timing of any agent-led surge remain unknown.

Public Services Face Design Choices

The immediate test for public bodies is how to manage requests under existing rules while keeping access open. The ICO guidance gives authorities a basis for handling AI-written FOI submissions: treat them as valid, and ask for clarification where needed. The Sky News report does not identify a new government policy or timetable for a wider response.

Over time, services may face pressure to replace free-form written applications with structured forms and clearer digital processes. Loosemore warns that making submissions harder could prompt technology providers to work around those barriers, while excluding people who need help expressing a claim. Whether departments redesign application and complaints systems, and how they will protect fair access while managing workload, remain open questions.

Key Questions

What did the Sky News report find?

It reported rising complaints, FOI requests and employment tribunal claims across several UK services since ChatGPT launched in November 2022. The figures vary by channel, and the report does not prove AI caused the increases.

Does the report show that most complaints are AI-generated?

No. A small test of 100 recent FOI requests found that an AI detector labeled 89% fully AI-written, but the report warns that detectors are unreliable. The test does not establish authorship across all requests or prove what caused rising volumes.

Are the additional complaints mostly frivolous?

The report does not support that conclusion. It cites researcher Chris Schmitz’s view that most cases in his research involve people seeking something they are entitled to claim. It also reports that a lower share of DWP complaints is upheld, while noting that the larger total may still mean more upheld complaints in absolute terms.

What guidance applies to AI-written FOI requests?

The Information Commissioner’s Office says an AI-written FOI request remains valid. Public authorities do not need to establish who or what wrote it and may ask for clarification if the request is unclear or badly formed.

What might change as AI agents become more common?

Agents could act for users by submitting requests or claims, rather than only drafting text. The report raises this as a possible next development, but the extent and timing of any resulting increase in submissions are not yet known.

Source: ThorstenMeyerAI.com