What Happens When AI Floods Britain’s Offices With Paperwork?

🔍 Read the full analysis: What Happens When AI Floods Britain’s Offices With Paperwork? on ThorstenMeyerAI.com

Age 18–24?Offer from Amazon

Prime made for students and young adults

• Fast, free delivery for dorm and study essentials
• Prime Video and Amazon Music included
• Member-only deals

Try Prime for Young Adults →Free trial for eligible 18–24 year olds

As an affiliate, we earn on qualifying purchases.Hide

A Sky News report published October 20 found that written complaints and claims across several UK public services increased after ChatGPT launched in November 2022, with 58% of surveyed civil servants saying AI-generated material from the public significantly affects their work. The figures show a broad rise in paperwork, but do not establish that AI alone caused it; the report also cites evidence that many people are using the tools to pursue claims they may be entitled to make.

Using data reviewed by Sky News technology correspondent Rowland Manthorpe, the report describes increases across multiple 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 rose 71%. Complaints to the Department for Work and Pensions increased 64%, while individual employment tribunal claims rose 62%. The report presents these as changes since late 2022; the figures do not, by themselves, show what share of each increase was caused by AI.

The increases were not limited to those services. Sky News reported rises in complaints about legal services, the NHS, police forces and public services. Each channel has its own possible drivers: water companies, for example, have faced difficult years. The report’s central observation is that several otherwise different categories had been relatively flat before rising after late 2022. It does not establish that the timing proves a causal link.

Manthorpe also tested 100 recent freedom-of-information requests with an AI-text checker. The tool labelled 89% as fully AI-written and 9% as human; a smaller 2019 sample was labelled entirely human. Manthorpe cautioned that the test was not conclusive: AI detectors can be unreliable, and the results do not prove that AI produced the wider rise. The Information Commissioner’s Office said in May that AI-written FOI requests remain valid, while authorities can seek clarification when a request is unclear.

More Claims, More Work for Public Bodies

For public bodies, the immediate issue is capacity: more letters, complaints and claims can take staff time to assess, clarify and answer. The civil-servant survey cited by Sky News indicates that many officials already feel a substantial effect. One respondent described being “bombarded with correspondence at speed”; others said some exchanges felt like arguing with ChatGPT and could be exhausting or anxiety-inducing.

But volume is not the same as abuse. At the DWP, Sky News reported that the share of complaints upheld fell from somewhere in the 50% range to close to 40%. That is a lower proportion, not evidence that most complaints are baseless. Because the total number of complaints grew, the report says, the absolute number of upheld cases may still be higher. The figures suggest policymakers must weigh staff workload against the possibility that easier writing tools help people bring legitimate concerns forward.

The potential next step is AI agents that can submit or pursue claims, rather than simply draft text for a person to send. If routine actions such as challenging a parking ticket or claiming compensation for a delayed train become automated, existing processes may face many more submissions at once. How much this happens, and how services will respond, remain open questions.

From ChatGPT to Automated Claims

The reported pattern begins after ChatGPT’s November 2022 launch, but the underlying channels are not new. People have long complained to public authorities, submitted FOI requests and brought employment claims. What has changed is that generative AI can reduce the effort involved in drafting formal, detailed correspondence. Sky News says much of today’s activity still involves people copying text between a chatbot and a service’s form or email.

The issue has also been recorded outside Britain. A paper by researcher Chris Schmitz, reported by TechCrunch in September, identified 84 possible cases of what he termed “agentic flooding” across 11 jurisdictions. The source material says many cases were roughly flat before 2022 and grew afterward. It also cites the UK Housing Ombudsman’s rise from about 2,600 complaints in 2022 to just over 7,000 last year, and a fivefold increase in complaints to the US Consumer Financial Protection Bureau. These examples show a wider pattern, not proof that AI explains every increase.

Some institutions have tried adding friction. The Financial Ombudsman Service found that about a third of its complaints were AI-generated and began charging some professional representatives to bring complaints; the number of complaints then fell. That example illustrates one possible response, though it does not settle whether fees or other barriers would screen out invalid cases or deter people with valid claims.

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

— Civil servant quoted in the survey cited by Sky News

Cause, Validity and Scale Remain Open

The reported figures establish increases in several forms of written contact, but they do not isolate the effect of AI from other possible causes, including changes in service quality, public awareness, policy or the number of people eligible to complain. The AI-checker test is a small sample and relies on a tool Manthorpe says is unreliable; it cannot establish who wrote each request or why the overall totals changed.

It is also unclear how many of the additional submissions are valid, how much time they add to each agency’s workload, and whether automated agents will substantially increase volumes. The DWP figures cited in the report show a lower share of complaints upheld, but do not provide a complete account of the merits of every submission. Schmitz’s findings and the reported upheld complaints point to genuine demand as well as administrative strain; neither establishes a single explanation for the trend.

Public Services Face a Design Choice

The immediate next step for authorities is to handle AI-written requests under existing rules. The Information Commissioner’s Office guidance says public bodies need not determine whether a request was written by a person or AI; they can ask for clarification if it is poorly formed. The source material does not identify a new nationwide policy or timetable for changing those rules.

Loosemore’s proposed alternative is to move services away from essay-like applications and toward structured data that captures the information needed to decide a case. He argues that adding obstacles is a poor long-term response because technology providers may find ways around them, while barriers could also deter people with legitimate claims. Whether government agencies adopt that approach, and how they protect access while managing workload, has not been settled.

The prospect of agents filing claims automatically makes that design question more pressing, but the scale and timing of such use remain uncertain. For now, the evidence describes a rise in written submissions and reported staff impact; the next policy choices will determine whether public services respond by restricting access or by changing how they receive and process requests.

Key Questions

What did the Sky News report find?

It reported that complaints, FOI requests and tribunal claims increased across several UK channels after ChatGPT launched in November 2022. The increases vary by service and do not prove that AI caused them.

Does the evidence show that AI caused the increase?

No. The report describes a timing pattern and a small AI-detector test, but says the test is not conclusive. Other factors may also have contributed to rising complaint and claim totals.

Are AI-written FOI requests valid?

The Information Commissioner’s Office guidance cited by Sky News says AI-written requests are still valid. Authorities do not need to identify whether a person or AI wrote them, and can seek clarification if a request is unclear.

Are the additional complaints mostly false?

The report does not establish that. It cites a lower share of DWP complaints being upheld, while noting that the larger total may still mean more upheld complaints in absolute numbers. Researcher Chris Schmitz said most cases in his review involved people claiming something they were entitled to claim.

What changes could public services make?

Tom Loosemore argues for replacing essay-heavy processes with structured data and redesigned services. No government-wide change or timetable is identified in the source material.

Source: ThorstenMeyerAI.com