AI ‘Clog’ Plagues UK Public Bodies Amid Surge in Chatbot-Generated Complaints

“It started as a trickle in 2024 and now it’s an onslaught.” Hollie Wickens, a senior caseworker within a Labour MP’s office, describes the daily reality of managing a surge of artificial intelligence-generated correspondence.

“Half the emails we receive are some form of AI,” she says, adding that this has made requests and complaints considerably more complex and time-consuming to process, despite her decade of experience in the position. “It is definitely more taxing.”

At the beginning of the 2020s, applying for employment, submitting complaints, or objecting to local planning decisions typically required substantial effort: research, form completion, and formal correspondence—often demanding specialized knowledge.

That shift began with the rapid expansion of ChatGPT – now four years old later this month – alongside Claude, Copilot, Gemini and other generative AI models. These tools empower non-specialists by providing access to free yet potentially unreliable legal, administrative, or personal assistant capabilities.

AI tool developers initially forecast substantial productivity gains for employers and their teams, arguing that routine tasks could be automated rapidly, thereby freeing human employees for more creative endeavour. Signs of such efficiency appear across parts of the economy, with AI contributing to improved UK output in recent reports.

However, UK businesses, local authorities, courts, schools and other public bodies are reporting being swamped by a flood of AI-generated complaints, letters and legal challenges that massively increase their workload, already straining limited resources.

This phenomenon has been described as “AI clog” by commentators.

‘Everyone’s Got a Free Lawyer’

With the rise of low-quality “AI slop” now mainstream, organisations warn that operations are being undermined by ever-increasing volumes of correspondence.

These requests, though instant to create, demand significant human time to sift through and respond. Researchers from the Centre for Digital Governance at the Hertie School in Berlin, the Cooperative AI Foundation and GovAI have labelled AI-generated spams “agentic flooding.”

“Every sector is experiencing very similar issues,” says Deborah Evans, chief executive of Lawyers in Local Government (LLG), which represents 5,600 lawyers working in local authorities. “AI is undeniably a powerful tool that accelerates many tasks, though it can also make complaints sound overly formal and persuasive.”

Local authorities face surges in complaints spanning adult social care services, missed bin collections, and planning application objections. “Local authorities operate with one constant budget and receive no additional funds due to increased complaints,” Evans explains.

Officials note that the length and characteristic style of correspondence serve as indicators of artificial generation, such as excessive length, repetitive phrasing, and the inclusion of unrelated comments or misdirected demands.

“This is a genuine problem,” says councillor Alessandro Georgiou, leader of Enfield council. “Residents are increasingly using AI to appeal local decisions, resulting in a significant rise in parking appeals.”

The Information Commissioner’s Office (ICO) reports that public authorities received more requests in 2025, finding that AI-generated communications are often very long, repetitive, and may contain inaccurate details. Before the chatbot era, average complaints were about two pages; today, most are ten times longer, with one notably exceeding 200 pages.

The ICO cautions that requests resembling AI drafts with inaccuracies still require human review. AI tools can misinterpret legal requests, citing case law from inappropriate jurisdictions or occasionally fabricating information due to hallucination.

Certain employers worry that AI’s tendency toward over-validation can give complainants misplaced confidence to pursue legal action.

“Everyone has a free lawyer now,” complains the founder of one technology start-up. A former employee, let go on performance grounds, is challenging their dismissal through an employment tribunal, though the entrepreneur asserts the process was handled correctly. “The wording reveals AI generation, and the tone is distinctly different.”

Amid a backlog of employment tribunal cases, the founder expects to spend £30,000 on legal fees by the time the case concludes.

National employment tribunal data shows a 28% surge in claims opened from April to June compared to the previous year, representing 14,000 cases from single claimants. Concluded cases also fell, leaving 70,000 open single claims at the end of June.

This experience is not unique, according to Jim Moore, employee relations expert at Hamilton Nash, whose employers report receiving correspondence that is either entirely AI-generated or heavily reliant on AI wording.

Employers describe received complaints as “unnecessarily verbose” and filled with “embellished or exaggerated claims that portray minor incidents as major offences.”

Moore notes that many AI-generated complaints “miss out on the most important point of a complaint – what resolution is sought? Most are worded to be intimidating but fail to outline the specific remedy the complainant desires.”

Statutory response times for some requests are compounding pressure on clogged public bodies.

Transport for London (TfL), responsible for the capital’s public transport system, recorded a 5% rise in freedom of information (FoI) requests in 2025-26, representing a growth from the 2023-24 peak. The number of appeals has also increased, with a reported rise in requests perceived as AI-generated that contain “highly technical and often inaccurate assertions and assumptions.”

Westminster Inundated

Offices of some MPs are also struggling to cope.

Vonnie Sandlan, head of office for Graeme Downie, the Labour MP for Dunfermline and Dollar, says surges in AI-generated requests have left her staff “exhausted.” “They are stretched to the limits and working more hours than they are paid,” she reports.

She notes that AI agents could assist those who struggle to draft correspondence, yet her office has recently stopped responding to automated “click to email your MP” campaigns to prioritize constituent requests.

The GMB union, representing MPs’ office staff, has found soaring casework levels widespread. “AI is not lightening the load in MPs’ offices,” says Lisa Gillmore, president of the MPs’ and peers’ staff branch. “Where offices keep their own figures, that casework has doubled or more, and staff carry the burden without any let‑up.”

Evans warns that every pound spent by local authorities on handling complaints is a pound not available for other priorities, underscoring the need for control. Some ombudsman services have introduced word limits for complaints or designed online templates to restrict attachment sizes.

Others fear organisations might use AI to sift through AI-generated correspondence—human‑to‑machine review—but this approach lacks the “human touch,” as Evans puts it.

Wiltshire council has explored using AI tools to summarise lengthy emails and draft initial responses, while a committee at Stroud district council in Gloucestershire has discussed employing Microsoft Copilot to assist with report writing and information presentation.

Researchers have also suggested authorities could step in by making applications harder to complete, charging fees for submissions, or blocking bots from government websites.

The UK government has issued advice on AI tool use for civil servants, including verifying facts, avoiding personal data input, and ensuring humans review and make final decisions.

Technology journalist Jamie Bartlett has been approached by many sharing experiences of the “AI clog.” He warns that public bodies and companies will likely need some form of AI to filter out unwanted correspondence, or adopt stricter acceptance criteria, or risk being “completely overwhelmed.”

Source link