TL;DR

Complaints to UK public bodies are rising as people use AI chatbots to write them, The Independent reported on 25 October. Police forces, schools and councils describe the submissions as long, often wrong and padded with meaningless legal-sounding language. Caseworkers in MPs’ offices told the paper that roughly half their cases now involve AI in at least some of the drafting.

A heavier load on stretched services

Councillor Richard Clewer, who leads on public service reform and innovation for the Local Government Association, said the complaints are getting “longer and more complex”, which is “increasing administrative burdens at a time of significant resource pressure”. Councils are looking at what government action might be needed to keep their frameworks workable.

The paper also cited figures compiled by The Economist showing employment tribunal claims up 39% in the first quarter of this year. Some claimants now face waits until 2029.

Genuine grievances caught in the queue

Not every AI-assisted complaint is frivolous. Journalist Sam Freedman, who has written about the trend, said AI helps people who previously struggled to put a complaint into words. The problem, in his view, is collective: complaining gets easier for each individual, but backlogs grow for everyone. He warned bodies may respond by making complaints harder to lodge or by using AI to answer them, ending in “chatbots talking to each other”.

Hallucinations in the paperwork

Consumer campaigner Helen Dewdney, the blogger behind The Complaining Cow, said chatbots get “a huge amount wrong”. In her test, a broadband complaint threatened Ofcom rather than pointing to the right dispute resolution scheme, and in another case the tool referred to an ombudsman that does not exist.

Looking Forward

The worry extends to the courts. In a Mansion House speech on 22 October, FCA chief executive Nikhil Rathi said legal claims drafted with AI could add to delays in a judicial system already under strain. For UK organisations that handle complaints, in our view the practical answer is triage that checks claims against facts quickly, without assuming AI-drafted means unfounded. Dewdney’s advice to complainants is to collect accurate information from reliable sources first and use the chatbot only to assemble it.