AI Is Making Paperwork Harder To Escape In Britain’s Bureaucracy

🔍 Read the full analysis: AI Is Making Paperwork Harder To Escape In Britain’s Bureaucracy on ThorstenMeyerAI.com

Complaints, information requests and tribunal claims to British public bodies have increased across several channels since ChatGPT launched, according to a Sky News report published October 20. The report says 58% of surveyed civil servants considered AI-generated material from the public a significant influence on their work, while stressing that the data do not prove AI caused the growth.

Sky News technology correspondent Rowland Manthorpe compared figures for a range of public-facing processes. Since ChatGPT’s launch in November 2022, household complaints about water companies rose 144%, complaints to English local authorities increased 78%, and written complaints to GP practices climbed 71%. The report also records increases in complaints to the Department for Work and Pensions, employment tribunal claims, legal-services complaints, NHS complaints, police complaints and complaints about public services.

The increases vary by channel, and each has its own possible causes. Water companies, for example, have faced a difficult period. Sky News argues that the repeated pattern—a long period of relatively little movement followed by a rise after late 2022—is consistent with AI making it easier to prepare formal submissions, but it is not proof of a direct cause.

Manthorpe also tested 100 recent Freedom of Information requests with an AI-text checker. It classified 89% as fully AI-written and 9% as human-written. A smaller 2019 sample was classified as entirely human. The report cautions that AI detectors are unreliable and that this limited test cannot establish how much AI contributed to the wider rise.

More Claims, Same Paperwork Systems

The increase matters because many UK public services still ask people to explain problems or make claims through written forms, letters and appeals. If AI lowers the time and confidence needed to produce those submissions, agencies may receive more cases without having more staff or redesigned systems to process them. Civil servants surveyed earlier this year described the impact in direct terms: 58% said AI-generated public submissions significantly affected their work.

Higher volume does not by itself mean a wave of fabricated complaints. Sky News reports that the share of DWP complaints upheld has fallen from the 50% range to close to 40%, but a lower proportion applied to a much larger number can still mean more complaints upheld in total. Researcher Chris Schmitz, whose work on possible cases of automated mass submissions was reported by TechCrunch, said: “The vast majority of cases we find are people who are entitled to claim for something, claiming for that thing.” That finding suggests some of the pressure may reflect previously unmet demand rather than artificial claims.

For readers, the stakes include how easily they can challenge a decision, seek compensation or ask for public information. Restricting AI-written submissions might reduce workload, but could also restore barriers for people who have a valid claim and struggle to write formal correspondence. The report frames the policy choice as a question of whether to add friction or redesign services around clearer, structured information.

Guidance and a Growing Trend

In May, the Information Commissioner’s Office issued guidance on AI-generated Freedom of Information requests after authorities reported higher volumes and more complex submissions. Some requests, the regulator said, misquoted legislation or needed substantial clarification. The ICO’s guidance says an AI-written request remains valid; public authorities do not need to identify whether a person or an AI produced it, and they may seek clarification when a request is unclear.

The pattern is not limited to Britain. A paper by Chris Schmitz, covered by TechCrunch in September, describes 84 possible cases of “agentic flooding” across 11 jurisdictions, with many rising after 2022. The UK Housing Ombudsman received about 2,600 complaints in 2022 and just over 7,000 last year, according to the source report. The US Consumer Financial Protection Bureau’s complaints grew fivefold. These examples point to a wider administrative challenge, though the cases and institutions are not directly interchangeable.

So far, much AI assistance involves a person using a chatbot to draft or improve a letter. The report also demonstrates an AI agent that can take actions on a user’s behalf. That shift could make it easier to submit applications, appeals or compensation requests repeatedly, without drafting each one by hand.

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

— Civil servants surveyed in research cited by Sky News

What the Data Cannot Prove

The reported increases show a pattern, but they do not isolate AI as the cause. Changes in service quality, public awareness, eligibility, backlogs or other events could also affect complaint and claim numbers. The source material does not provide enough detail to compare every channel’s figures on a common basis or to establish how much of each increase came from AI-assisted submissions.

The FOI text-checker exercise is also limited: it involved 100 recent requests and a smaller 2019 sample, and AI-detection tools can misclassify writing. It cannot establish that 89% of all current FOI requests are AI-written. Nor does a rise in submissions establish that each one is valid. The report’s evidence points to a mixture of increased administrative pressure and genuine claims, but the exact balance remains uncertain.

It is also unclear how widely AI agents will be used to file claims automatically, and whether public agencies can distinguish useful automation from repetitive or misleading submissions without blocking people with legitimate requests.

Redesigning Applications and Appeals

The immediate response from the ICO is to treat AI-written FOI requests as valid and ask for clarification when needed, rather than requiring authorities to identify their authors. Other agencies will have to decide how to manage increased correspondence while maintaining access to complaints and appeals.

Loosemore’s proposed alternative is to replace essay-like applications with structured data and simpler digital services. That approach would ask people for specific information in a consistent format, rather than relying on long written explanations. The report does not set out a government plan or timetable for such a change. The next test will be whether agencies can update their processes before AI agents make routine submissions even easier to file.

Key Questions

What has increased in Britain since ChatGPT launched?

Sky News reports increases across several channels, including household complaints about water companies, local-authority and GP complaints, DWP complaints, Freedom of Information requests and individual employment tribunal claims. The reported growth differs by channel.

Does the report prove AI caused the rise?

No. The timing and repeated pattern are consistent with AI making written submissions easier, but the report does not establish causation. Other changes may also have affected complaint and claim volumes.

Are AI-written FOI requests valid?

The Information Commissioner’s Office says AI-written requests remain valid. Authorities do not need to determine who or what wrote a request, though they can ask for clarification if it is unclear.

Are the added complaints mostly fake?

The available evidence does not show that. The report cites research suggesting many people are claiming something they are entitled to, while also noting that the share of DWP complaints upheld has declined. The exact proportion of legitimate claims across all channels is not established.

What changes could public services make?

Tom Loosemore argues for services that gather structured information rather than relying on lengthy written explanations. No timetable or government-wide plan for that change is provided in the report.

Source: ThorstenMeyerAI.com