When AI increases complaints, the answer is better service - not higher barriers
Updated: Sep 22
A rise in AI-assisted complaints and applications should not automatically be treated as a problem to suppress. It may be evidence that people are finally able to navigate systems that were previously too complex, time-consuming or intimidating to use.
A recent TechCrunch report stated that public bodies in several countries are seeing sharp increases in requests as generative AI makes it easier for people to understand correspondence, complete forms and pursue complaints. In the UK, complaints to the Housing Ombudsman reportedly rose from 2,600 in 2022 to more than 7,000 last year; the article also notes similar growth in consumer, judicial and parliamentary channels internationally.
For public services, this creates an urgent operational challenge. But it also presents an opportunity: use AI not simply to process a greater volume of complaints, but to prevent avoidable complaints by helping people get what they need earlier.
The signal behind the volume
It is tempting to describe a rapid growth in complaints as “AI-generated noise”. Certainly, public bodies will need safeguards against duplicated, malicious or low-quality submissions. But that framing risks missing the more important point.
Many people do not complain, appeal or apply for support because they do not have a legitimate issue. They do not do so because the route is hard to find, the process is confusing, the language is inaccessible, or the effort required feels disproportionate to the likely outcome.
A resident may receive a housing-repairs update full of jargon and have no idea what happens next. A person seeking care support may not understand what they are eligible for, what evidence is required, or how to challenge a decision. Someone dissatisfied with a council service may give up rather than work out which team is responsible and how to describe the issue in the right format.
AI reduces that burden. It can explain a letter in plain English, turn a timeline into a coherent complaint, identify missing information, translate content, and help someone ask the right question.
That is valuable. But if public organisations only encounter AI after it has helped a frustrated resident submit a formal complaint, they are intervening too late.
Shift from complaint handling to need resolution
The strategic question should not be: “How do we cope with more AI-generated requests?”
It should be: “How do we use AI to resolve the underlying need before a formal complaint becomes necessary?”
This is the distinction between an AI-enabled complaints function and an AI-enabled service.
A well-designed public-service assistant can support residents at the earliest point of uncertainty. It can help people understand a decision, find the correct service, check the status of a request, prepare the right evidence, receive a clear explanation of policy, and access a human adviser when the matter needs empathy, discretion or specialist judgement.
In practice, that means building AI into the front door of services - not as a barrier, but as an accessible guide.
Traditional response | Better AI-enabled response |
Receive a complaint after trust has already broken down | Identify the person’s need and offer a clear path to resolution early |
Ask residents to navigate service silos | Guide them to the correct team, form, entitlement or next step |
Send standard acknowledgements | Provide tailored, understandable explanations and realistic timescales |
Treat repeat contact as a volume problem | Use contact patterns to identify process failures and unmet needs |
Focus on closing cases quickly | Focus on achieving a fair, evidenced and sustainable outcome |
Escalate only after a resident has struggled | Offer proactive support when a request appears stuck or incomplete |
The purpose is not to make complaints harder to submit. Formal complaint routes must remain accessible, independent and fair. The purpose is to reduce the number of complaints that arise because a person could not get a straightforward answer, meaningful update or reasonable resolution through normal service channels.
What this could look like
Consider a tenant reporting a persistent repair issue.
Without effective digital support, they may submit a repair request, receive an unclear update, wait beyond an expected timescale, call several times, search unsuccessfully for escalation information, and eventually use an AI tool to draft a detailed complaint.
The complaint may be entirely justified. Yet much of the frustration could have been avoided if the service had offered:
A plain-language explanation of the repair category, target timescale and current status.
A clear account of why a delay has occurred and what will happen next.
A simple way to upload photographs, clarify the impact on the household and request a reassessment where circumstances have changed.
Proactive contact when an appointment is missed, an issue remains unresolved or a case exceeds its expected timeframe.
An AI assistant that can answer service-specific questions using approved, current information rather than generic web content which may be out of date.
A straightforward route to a trained person when the issue involves risk, vulnerability, dissatisfaction or a need for discretion.
In that scenario, AI is not being used to deflect the tenant. It is being used to give them clarity, agency and a reliable route to a solution.
That is what good public-service technology should do.
How Datnexa helps
At Datnexa, we work with public-sector organisations to make AI practical, safe and genuinely useful for the people who rely on their services.
Our focus is not on deploying AI for its own sake, or using automation to create another barrier between residents and support. It is on helping organisations design and implement AI-enabled services that make information easier to understand, journeys easier to complete and outcomes easier to achieve.
That can include:
Developing secure, organisation-specific AI assistants that use approved policies, procedures, service information and knowledge content.
Improving resident and service-user journeys by helping people find the right service, understand eligibility, prepare information and identify their next step.
Supporting frontline teams with faster access to accurate guidance, case summaries and consistent communications.
Using contact, enquiry and complaint themes to identify recurring pain points, unclear processes and opportunities for service improvement.
Embedding appropriate governance, privacy, transparency, human oversight and escalation routes from the outset.
Helping teams build the skills, confidence and operating models needed to use AI responsibly in day-to-day public services.
For councils, housing providers, adult social care organisations and other public bodies, the opportunity is to use AI as a layer of support around existing services: helping people navigate complexity while enabling staff to spend more time on cases where human expertise and judgement matter most.
The result should not simply be a more efficient complaints process. It should be a better experience before a complaint is ever needed.
Design AI around trust
There is a real risk that organisations respond to rising volumes by placing more automation between people and help. This would be a mistake.
Residents can quickly tell the difference between technology that helps them and technology that contains them. An AI system that offers generic answers, conceals escalation routes, makes decisions without explanation or sends people in circles will create more dissatisfaction - not less.
Public bodies should therefore design AI around a set of clear principles:
Solve before you screen. Start by understanding the person’s issue and helping them make progress, rather than treating every interaction as a request to be filtered.
Use trusted local knowledge. Responses should draw on approved service information, current policies, eligibility rules, live case status where appropriate, and clearly governed knowledge bases.
Make the next steps explicit. Residents should know what will happen, who owns the action, when they should expect an update and how to seek further help.
Protect human access. People must be able to reach a person, particularly where there is vulnerability, safeguarding risk, urgent need, language barriers or an unresolved dispute.
Be transparent about automation. Explain when AI is being used, what it can and cannot do, how personal information is handled, and how a decision can be reviewed.
Learn from demand. Analyse recurring questions, repeat contacts, abandoned journeys and complaint themes to improve the underlying service - not merely the assistant’s wording.
Preserve accountability. AI can support triage, drafting, summarisation and insight, but organisations remain responsible for the outcomes their residents experience.
A practical opportunity for public services
The growth of AI-assisted complaints should be understood as a warning and an invitation.
It is a warning that administrative friction no longer protects poor processes from scrutiny. As AI helps more people understand their rights and express their concerns, organisations may see demand that was always present but previously hidden by complexity.
But it is also an invitation to rethink how services are designed.
The most effective use of AI in public services will not be an automated system that simply handles more complaints at lower cost. It will be an integrated support model that helps people get accurate information, complete the right task, understand what is happening, and access help before their issue becomes an escalation.
That means joining up customer service, case management, knowledge management, complaints insight and operational improvement. It means treating every complaint not only as a case to close, but as evidence of a service journey that may need redesigning.
AI can make it easier for people to complain. Public services should respond by making it easier for people to be heard, understood and helped.
The goal is not fewer complaints at any cost. The goal is fewer avoidable complaints because people received the support, clarity and resolution they needed in the first place.
