Surge in AI-assisted complaints is putting extra strain on councils and schools, BBC investigation finds

[LokiCurrent]

A BBC investigation has found that a rising tide of complaints written with AI assistance is substantially increasing the workload of schools, local councils, and other UK public bodies. Chatbots can produce complex, legally styled arguments that quote regulations or acts of Parliament directly, and one public body told the BBC that complaints which used to fit on a single sheet of A4 paper can now run to twenty pages long once AI gets involved in drafting them.

The scale of the added workload is significant according to those actually handling it. Deborah Evans, chief executive of Lawyers in Local Government, which represents 5,600 lawyers working across local authorities, estimated her organization would need double its current workforce just to keep up with the increase in complaint volume and complexity. AI generated complaints also frequently contain errors, sometimes misquoting the actual legislation they cite, which adds even more time to the process since staff have to verify claims that sound authoritative but turn out to be wrong.

Schools specifically report a similar pattern. Pepe Di'Iasio, general secretary of the Association of School and College Leaders, told the BBC that while complaints are rising for a variety of reasons, AI is adding meaningfully to the effort required to resolve them, since parents are increasingly using AI to generate multiple, extensive complaints that put further strain on already stretched staff. The Local Government and Social Care Ombudsman said AI needs to be used with discretion, acknowledging the difficulty of preserving the genuine accessibility benefits AI offers while discouraging complaints that mostly just waste institutional time.

Not everyone experiencing this shift sees it negatively. Several people told the BBC that AI has made complaining more accessible and effective, with one person specifically describing it as having made a real practical difference in how they've been able to represent themselves against an institution that previously held most of the procedural advantage. That tension, between AI actually lowering the barrier for individuals to advocate for themselves and the same technology overwhelming the institutions meant to respond, is really the core of the story, and nobody quoted in the piece has a clean answer for how to preserve one benefit without the other cost

Surge in AI-assisted complaints putting extra strain on councils and schoolswww.yahoo.com

[James_46]

The errors and misquoted legislation detail is the part that actually worries me most reading through this. A confidently wrong AI generated complaint citing law that doesn't actually say what it claims still has to get investigated and formally responded to by an overworked human on the other end. That's pure added friction with essentially zero corresponding value for anyone involved on either side of the exchange

[NatureBoyMichael27]

Curious whether any of these public bodies are actually exploring using AI on their own end specifically to help triage and respond to this rising volume more efficiently. Feels like the obvious next move given the scale of the problem described here, fight AI generated volume with AI assisted response capacity rather than trying to hold the line entirely with existing human staffing levels alone

[NoLimitsBen18]

Wonder what percentage of these AI assisted complaints actually turn out to be legitimate, valid grievances that simply got expressed more effectively, versus ones that are mostly just noise or clearly vexatious complaints dressed up in more convincing legal sounding language. That distinction seems pretty important for figuring out what an actually sensible policy response should even look like here

[Velvet Connor]

This feels like a preview of a much broader pattern that's going to keep showing up across every institution that has some kind of formal complaints or appeals process, not just UK schools and councils specifically. Insurance disputes, planning permission appeals, benefits tribunals, any bureaucratic process with a written submission component is going to see the exact same dynamic play out eventually.

Institutions built their staffing and process assumptions around a certain average complexity and volume of complaints arriving through entirely human effort. AI breaks that baseline assumption completely, and most of these systems have no obvious way to scale their actual response capacity anywhere near as quickly as the input volume and complexity itself can scale up

[Memory Jamie]

The point about AI making complaining more accessible for people who previously had no real way to push back against an institution is honestly the part of this story that deserves more sympathy than it usually gets in coverage focused mainly on the burden side. Plenty of people properly don't know how to write a formal complaint that references the right regulations or legal frameworks, and AI closing that literacy and confidence gap is a real, meaningful benefit for people navigating bureaucracy that was never built with them specifically in mind.

The problem isn't really that people are complaining more effectively now, it's that the volume and error rate scaled up faster than institutions could adapt their own processes to handle it. Those are two notably separate problems that unfortunately compound each other in exactly this specific situation

[Badger27]

Needing double the current workforce just to handle the increase is a genuinely staggering number for an organization to state publicly. That's not a minor efficiency drag, that's a fundamental capacity crisis building in slow motion across the whole sector

[Slow Tiger]

Twenty page complaints replacing single sheet ones is such a vivid, concrete way to actually illustrate this problem. Numbers alone rarely land the same way that specific example does

[ElectronCloud]

The discretion advice from the Ombudsman is basically just asking people to self regulate their own AI use responsibly, which rarely actually works at scale once a genuinely useful tool like this becomes widely available and known about. Hard to see this getting meaningfully better without some kind of actual structural or procedural change on the institutional side specifically, rather than relying purely on individual complainants choosing restraint on their own