Generative artificial intelligence has made it effortless for citizens to draft formal letters, dispute parking penalties, and challenge council rulings. However, public workers report that these automated submissions are overwhelming administrative offices across the United Kingdom. Staff must spend hours sifting through lengthy, repetitive documents generated in seconds, creating severe backlogs that delay responses for everyone.
In parliamentary offices, the influx has transformed routine correspondence. Hollie Wickens, a senior caseworker for a Labour MP, estimated that roughly half of the emails her office receives now involve artificial intelligence, making them far more complex and taxing to review. Researchers from the Hertie School, the Cooperative AI Foundation, and GovAI have described the bombardment as "agentic flooding," while observers have dubbed the administrative paralysis "AI clog." The surge affects local councils, schools, and courts at a time when their budgets are already under pressure.
Local authorities face particular strain from the ballooning length of submissions. According to Deborah Evans, chief executive of Lawyers in Local Government, which represents 5,600 municipal lawyers, the average complaint once spanned around two pages; the majority are now ten times longer, with the most extreme reaching 200 pages. Automated tools frequently treat routine complaints as formal lawsuits, citing irrelevant foreign statutes or hallucinating legal precedents. Despite these flaws, the Information Commissioner’s Office has advised that authorities must still assign human staff to answer them.
The disruption is also visible in the justice system, where employment tribunal claims opened between April and June jumped 28 percent compared to the previous year, reaching 14,000 single cases. Employers and administrators warn that chatbots, functioning like free and uncritical legal advisers, often give claimants false confidence to pursue weak disputes. While developers promised that generative tools would save time by automating routine paperwork, public servants increasingly find themselves trapped in managing the synthetic documents those tools produce.