How Artificial Intelligence Correspondence Is Overwhelming British Public Services
London, Sunday, 8 November 2026.
Surging volumes of complex, automated complaints—some expanding tenfold to 200 pages—are overwhelming British public institutions, doubling staff workloads and straining fixed taxpayer budgets with unprecedented administrative gridlock.
The AI Clog Phenomenon
British public institutions are grappling with an unprecedented administrative crisis termed ‘AI clog’ or ‘agentic flooding,’ where surging volumes of AI-generated complaints, legal challenges, and Freedom of Information (FoI) requests are overwhelming administrative systems. This phenomenon, first identified as a trickle in 2024, has escalated into a full-blown onslaught by late 2026, forcing public bodies to divert critical resources toward parsing and responding to synthetic correspondence rather than addressing genuine citizen concerns [1]. The issue spans across multiple sectors, including parliamentary offices, local authorities, legal bodies, and educational institutions, creating a uniform strain on public administration [1].
Mechanisms Behind the Surge
The democratization of generative AI tools such as ChatGPT, Claude, Copilot, and Gemini has enabled individuals to draft complex, formal complaints and appeals with minimal effort. Researchers at the Hertie School in Berlin, the Cooperative AI Foundation, and GovAI have labeled this phenomenon ‘agentic flooding,’ describing it as a deliberate or incidental strategy where AI-generated correspondence bombards public bodies [1]. The result is a staggering increase in the volume and complexity of administrative workloads. For instance, average complaint lengths for local authorities have ballooned from approximately two pages to ten times longer, with some complaints stretching to 200 pages [1]. This inflation in document length not only consumes time but also exacerbates the cognitive load on staff tasked with addressing these issues.
Operational and Financial Strain
The administrative strain is further compounded by the financial constraints of public institutions. Deborah Evans, chief executive of Lawyers in Local Government (LLG), which represents 5,600 lawyers working in local authorities, noted that ‘local authorities have one pot of money and don’t get extra money because of more complaints’ [1]. This means that resources allocated for genuine citizen services are increasingly diverted to handle AI-generated correspondence, creating a zero-sum scenario where addressing synthetic complaints directly impacts the ability to deliver essential public services. Transport for London (TfL) experienced a record 4,769 FoI requests in 2025-26, a 5% increase from the previous peak in 2023-24, with a notable rise in AI-generated appeals [1].
Human Cost and Workforce Impact
The human toll of this administrative crisis is significant. Hollie Wickens, a senior caseworker in a Labour MP’s office, revealed that ‘half the emails we receive are some form of AI’ and described the workload as ‘definitely more taxing’ [1]. Staff in MPs’ offices, including those representing constituencies such as Dunfermline and Dollar, are stretched to their limits, working additional hours without compensation. Vonnie Sandlan, head of office for Graeme Downie, the Labour MP for Dunfermline and Dollar, stated, ‘They are stretched to the limits and working more hours than they are paid’ [1]. Lisa Gillmore, president of the MPs’ and peers’ staff branch of the GMB union, corroborated this, noting that casework in MPs’ offices has doubled or more, yet staff continue to carry the burden without relief [1].
Quality and Accuracy Concerns
Beyond sheer volume, the quality and accuracy of AI-generated correspondence pose additional challenges. Organizations, including those in the public sector, are receiving longer, more complex, and sometimes inaccurate complaints [3]. This trend, highlighted by the Institute of Customer Service on X, underscores the need for human judgment and empathy in complaint resolution [3]. Jamie Bartlett, a technology journalist, emphasized that public bodies may need to deploy AI tools like Microsoft’s Copilot to filter and summarize correspondence, or implement stricter rules on acceptable submissions to mitigate the influx of low-quality, automated complaints [1].
Response Strategies and Future Implications
Public institutions are exploring a range of countermeasures to address the AI clog phenomenon. Proposed solutions include implementing word limits for complaints, creating standardized online templates, charging fees for excessive or frivolous requests, and blocking automated bots [1]. Some bodies are also turning to AI-assisted tools to streamline the processing of correspondence, though this raises questions about the long-term sustainability and effectiveness of such measures. The National Audit Office (NAO) is examining the role of AI in driving productivity gains within government, focusing on whether central authorities have established clear direction and enabling conditions for departments to integrate AI successfully [4]. However, the NAO’s current study does not extend to assessing AI adoption in local government or other wider public sector bodies, leaving a critical gap in understanding the full scope of the challenge [4].
Broader Regulatory and Societal Considerations
The rise of AI-generated correspondence also intersects with broader debates about AI regulation and transparency. The UK government’s 2024 ‘Growing Up in the Online World’ consultation revealed that young people’s most consistent demand was the mandatory labeling of AI-generated content, yet this demand was notably absent from subsequent government responses and House of Lords debates [5]. This oversight highlights a disconnect between public sentiment and policy action, which could further complicate efforts to manage the societal impacts of AI-generated administrative burdens.