How AI Is Changing The Volume Of Paperwork In Britain

🔍 Read the full analysis: How AI Is Changing The Volume Of Paperwork In Britain on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on monitors, keyboards and dev gear

• Business-only prices and quantity discounts
• Tax-exempt purchasing
• Multiple users, one account, clear invoices

Create a free Business account →Free to register

As an affiliate, we earn on qualifying purchases.Hide

A Sky News report published October 20 found that complaints, Freedom of Information requests and employment tribunal claims have risen across multiple UK services since ChatGPT launched in November 2022. The figures show a similar upward pattern across different channels, but do not establish that AI caused the increases; evidence cited by the report indicates that many complaints may reflect genuine, previously unmet demand.

Sky News technology correspondent Rowland Manthorpe examined data from complaints to water companies, councils, GP practices, the Department for Work and Pensions, and other public services, alongside individual employment tribunal claims. The report found growth ranging from 23% to 144% across the listed channels since ChatGPT’s launch. Household complaints about water companies rose 144%, complaints about English local authorities 78%, and written complaints to GP practices 71%. The report does not present these figures as proof that AI was responsible.

Other increases in the report include 64% for DWP complaints, 62% for individual employment tribunal claims, 49% for complaints about legal services, and 35% for written complaints to the NHS. Complaints about police forces rose 34%, while complaints about public services increased 23%. Each channel has its own possible causes: the report notes, for example, that water companies have faced difficult years. Its central observation is that several unrelated measures were relatively flat before turning upward after late 2022.

Manthorpe also tested 100 recent FOI requests with an AI-text detector. It classified 89% as fully AI-written and 9% as human; a smaller 2019 sample was classified as entirely human. The report cautions that AI detectors are unreliable and that this limited test cannot show whether AI caused the broader rise. A survey of civil servants cited by Sky News found 58% said public AI-generated content significantly affects their work.

Public Services Face More Written Claims

More submissions can mean greater pressure on public-service teams, which must read, assess and answer complaints, appeals and information requests. But a higher workload is not, by itself, evidence that the submissions are frivolous. If AI helps people explain a problem or complete a formal process, the added volume may include claims that would otherwise have gone unmade.

The report cites DWP data showing that the proportion of complaints upheld has fallen from the 50% range to close to 40%. That is a lower success rate, but the report says the much larger overall volume still means more upheld complaints in absolute terms than before. The figures point to a difficult policy choice: adding barriers may reduce workloads, but could also discourage people with valid claims from seeking redress.

The change may also expose a mismatch between services designed around written explanations and tools that make producing those explanations faster. Former Government Digital Service head Tom Loosemore argues that public services should rely more on structured information than essays. That approach would address the design of the process rather than trying to restrict who can use writing tools.

How AI Entered the Paperwork Flow

Generative AI tools can draft or refine letters, complaints and FOI requests. The Sky News report describes much of today’s use as people copying text between a chatbot and an online form. That can lower the effort involved in writing a formal submission, particularly for someone unsure how to phrase a request or appeal. It does not mean every increase in complaints is caused by AI, or that every AI-assisted submission is valid.

In May, the Information Commissioner’s Office issued guidance after public authorities reported higher volumes and more complex AI-generated FOI requests. Some requests, the ICO said, misquoted legislation or needed substantial clarification. Its guidance says AI-written requests remain valid, authorities do not need to determine whether a person or AI wrote them, and officials may ask for clarification when a request is poorly formed.

The report also describes a wider pattern beyond Britain. Researcher Chris Schmitz, in work reported by TechCrunch in September, identified 84 possible cases of what he terms “agentic flooding” across 11 jurisdictions. The report says many cases rose after 2022, including complaints to the UK Housing Ombudsman, which went from about 2,600 in 2022 to just over 7,000 last year. These examples add context, but do not establish a single cause for the changes.

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

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

Cause and Claim Quality Remain Unsettled

The reported timing does not prove that ChatGPT or other AI tools caused the increases. The figures cover channels with different rules and circumstances, and the report notes that some have independent reasons for rising complaints. The detector test involved only 100 recent FOI requests, used a tool known to be unreliable, and cannot establish how much AI contributed to the overall trend.

It is also unclear how much of the new correspondence is written by AI, edited with AI, or produced without it. The civil-servant survey measures reported effects on staff, not the share of all submissions created with AI. The figures provided do not settle whether the pattern will continue, how much extra work it creates in each service, or whether reforms to forms and processes can absorb the volume without delaying responses.

There is evidence that some submissions identify legitimate problems, but the available figures do not provide a full assessment of claim quality across every service. DWP complaint outcomes are one measure, not a verdict on all complaints, FOI requests or tribunal claims. The balance between valid demand and low-quality submissions therefore remains different across services and not fully quantified.

Services Weigh Process Changes

Public bodies will continue to process requests under existing rules, including the ICO’s guidance that AI-written FOI requests are valid and can be clarified when necessary. The report does not identify a single nationwide policy change or timetable for responding to the increase. Decisions about handling volume are likely to vary by service, depending on the kind of submissions received and the resources available.

A further pressure may come from AI agents that submit requests or claims, rather than simply drafting text for a person to send. Manthorpe’s report demonstrates an agent service from Meta and raises the possibility of automated appeals, such as challenges to parking tickets or claims for rail delay compensation. The scale and consequences of that use are not yet established.

Loosemore’s proposed direction is to replace essay-style applications with structured data and redesigned services. Whether government bodies adopt that approach, and whether it can reduce staff workload without obstructing valid claims, remains to be seen. The next evidence to watch is how complaint volumes, outcomes and processing times change as services respond to wider AI use.

Key Questions

Have complaints and requests risen since ChatGPT launched?

Sky News reports increases across several channels since November 2022, including water-company complaints, local-authority complaints, FOI requests and tribunal claims. The reported growth varies by category, and the figures do not prove AI caused it.

Does the report prove that AI caused the increase?

No. It identifies a similar upward pattern across multiple channels after ChatGPT’s launch, but timing alone cannot establish cause. Each channel may have other reasons for changing, and the AI-detector test described in the report is limited and not conclusive.

Are AI-written FOI requests valid in Britain?

According to guidance from the Information Commissioner’s Office, an AI-written request remains valid. Public authorities do not need to determine who or what drafted it, though they can ask for clarification if it is poorly formed.

Are the new complaints mostly frivolous?

The report does not find that they are mostly frivolous. It cites researcher Chris Schmitz’s assessment that most cases in his research involve people claiming something they are entitled to, while also noting that outcomes vary and the evidence is not a complete measure of every UK service.

What might public services change?

One proposal from former Government Digital Service head Tom Loosemore is to move away from processes that depend on long written explanations and use more structured information. The report does not say that a nationwide change has been adopted or set a timetable.

Source: ThorstenMeyerAI.com