AI-generated complaints flood courts, councils, and ombudsmen as "agentic flooding" strains public services
Published · updated · curated by AI Is Going Just Great
Source: aol.com ↗
"The difficulty is that AI rarely uses one word when a hundred will do. What might previously have been a short email complaining about bins not being emptied can arrive looking like a High Court legal action."
Complaints to the UK's Housing Ombudsman jumped from roughly 1,700 a year to around 7,000, and the Bar Standards Board logged a 25% rise in complaints over the past year — both blamed on AI. An academic study across 84 public bodies, from Britain to Japan to Germany, found that chatbots have slashed the cost of filing a complaint to near-zero, producing a surge in volume and complexity that human staff cannot process at anything close to the same speed.
The problem compounds itself. Officials send a reply; the complainant pastes it into ChatGPT and asks for flaws, challenges, and escalations; the next submission arrives longer and more combative than the last. Many of the underlying complaints are also inaccurate or nonsensical, since people assume whatever the chatbot told them is correct. Some institutions have started fighting back: the California Management Review now charges $75–$150 per submission after manuscript volume tripled. The academics behind the study warn that such deterrents may actually increase AI use while discouraging real people from filing claims at all.