AI-Generated Complaints Put Pressure on UK Public Services

The growing use of Artificial Intelligence to write complaints to public bodies is creating a new challenge for investigators, housing associations and local authorities. While AI tools can help people explain complicated problems and communicate more confidently, they are also producing lengthy, highly technical submissions that can make it harder for staff to identify the actual issue and respond effectively.

The debate has exposed two competing concerns. Complaint investigators say excessive, repetitive AI-generated correspondence is increasing their workload and making meaningful communication more difficult. Others argue that AI is giving people who previously struggled to express themselves a stronger voice when challenging poor Public Services.

Three perspectives from people responding to a report about the growing number of chatbot-generated complaints in the UK highlight the complexity of the issue. A housing association investigator describes the pressure on frontline complaint teams, while transformation consultant Dexter Jandoo and Sheffield resident Pete Bibby argue that AI-assisted complaints should be recognised as a tool for improving access to justice and public accountability.

Housing Complaint Investigators Face an Increasing Workload

A complaint investigator working for a housing association says the nature of the job has changed dramatically within a few months because of the growing volume of AI-generated correspondence.

The investigator, who has worked in the organisation for nine years, says they are familiar with housing policies and legislation and remain committed to helping tenants resolve problems. However, reviewing hundreds of pages of lengthy complaints has made the work increasingly difficult.

The concern is not simply that tenants are using artificial intelligence. It is that some complaints generated with chatbots contain extensive technical language, legal references and repeated arguments that make it harder to identify what the tenant actually wants the organisation to address.

Complaint handling depends on understanding the specific problem, establishing what happened and determining what action can resolve it. When a submission contains large amounts of unnecessary material, investigators may have to spend additional time separating the central grievance from supporting explanations and repeated claims.

This can create pressure for staff who already have to manage several cases, communicate with tenants, examine records and work within established complaint procedures.

Why Human-Written Complaints Can Be Easier to Understand

The housing investigator argues that a complaint written directly by a tenant can sometimes be more useful than a lengthy AI-generated document, even if it contains spelling mistakes, informal language or emotional expressions.

The reason is that human-written complaints often communicate the problem in the tenant’s own words. They can make it easier for investigators to understand what happened, how the person was affected and what outcome they are seeking.

By contrast, AI-generated letters can sound highly formal and legally sophisticated without necessarily reflecting the complainant’s understanding of the language being used.

The investigator says that when staff speak to customers in person or over the telephone and use the same technical terminology found in their AI-assisted letters, some tenants struggle to understand the discussion.

This creates a communication gap. A document may appear professionally prepared, but the person who submitted it may not be familiar with the legal arguments, procedural language or technical expressions included in the text.

For complaint teams, the most useful information is often straightforward: what went wrong, when it happened, how it affected the tenant and what the organisation needs to do to put matters right.

The Awaab Ishak Case Highlights Why Accessible Complaints Matter

The investigator also referred to the case of Awaab Ishak to explain why housing complaints processes must remain accessible to tenants.

Awaab Ishak was a two-year-old boy who died in December 2020 after prolonged exposure to mould in the family’s social housing flat in Rochdale. An inquest in 2022 concluded that his death was caused by a severe respiratory condition resulting from exposure to mould.

The case exposed serious failures in how housing concerns were handled and became a major focus of public attention on conditions in social housing. It also demonstrated the consequences that can follow when complaints about unsafe living conditions are not properly addressed.

The investigator’s reference to the case underlines an important distinction: the ability to raise a complaint should never depend on a tenant’s education, writing skills or familiarity with legal terminology.

Tenants need to be able to report damp, mould, disrepair and other housing problems in ordinary language and receive an appropriate response. A complicated AI-generated letter should not be necessary to make an organisation take a serious complaint seriously.

The challenge for housing providers is therefore to manage additional correspondence without weakening the accessibility of their complaint systems.

Repeated AI-Generated Complaints Can Feel Like Harassment

The investigator also raised concerns about people who repeatedly submit large AI-generated complaints to the same organisation.

Frequent complainants are already familiar to many complaint-handling teams. However, the investigator says receiving multiple lengthy submissions every month from the same individual can make the workload particularly difficult.

Repeated correspondence can require staff to review earlier decisions, respond to overlapping allegations and distinguish new issues from matters that have already been addressed.

The investigator said the volume and nature of some submissions can make employees feel that their time and professional responsibilities are not being respected.

The housing association is considering whether certain patterns of repeated AI-generated correspondence could be treated as harassment. This raises a difficult question for public-facing organisations: how can they protect employees from abusive or unreasonable conduct without restricting legitimate complaints?

The distinction matters because frequent complaints are not automatically harassment. A tenant may have a genuine unresolved problem and may need to contact an organisation repeatedly when previous responses have failed to provide a solution.

Any policy addressing excessive correspondence would therefore need to consider the content, frequency and purpose of the communication rather than treating the use of AI itself as misconduct.

AI Can Help People Challenge Poor Public Services

Transformation consultant Dexter Jandoo offered a different interpretation of the growing number of AI-assisted complaints. He argued that using artificial intelligence to seek support or challenge failing public services should not automatically be considered a nuisance.

Jandoo described AI as a tool that can make it easier for ordinary people to pursue complaints that might previously have been difficult to prepare. Writing a detailed grievance can require time, confidence and an understanding of formal procedures, all of which may be difficult for people facing financial, personal or educational barriers.

AI tools can help users organise their experiences, structure arguments, identify relevant information and produce a more formal letter. This may allow people who would otherwise struggle to communicate with a council or housing provider to explain their concerns more effectively.

Jandoo argued that public bodies, particularly local councils, should focus on addressing the problems behind complaints rather than dismissing submissions because they were prepared with AI assistance.

He also supported the use of AI by public organisations themselves, provided that human staff remain involved in reviewing complaints and deciding how to respond.

This approach would involve using technology to manage large volumes of correspondence while retaining human judgement for issues that require investigation, interpretation and personal engagement.

AI Could Give a Voice to People Who Struggle to Write

Pete Bibby from Sheffield also challenged the idea that AI-generated complaints are inherently less legitimate than those written without technological assistance.

He argued that the use of artificial intelligence does not necessarily weaken a complaint or make its contents less worthy of attention. Instead, it may help people who have difficulty expressing themselves because of limited education, writing ability or confidence.

For some individuals, preparing a formal complaint can be intimidating. They may understand that something has gone wrong but struggle to explain the problem in language that a public organisation will recognise and respond to.

AI can help bridge that gap by turning a person’s account into a structured letter. It can organise events chronologically, improve readability and help users identify the main points they want to communicate.

Bibby suggested that the growing volume of AI-assisted complaints could reveal problems that previously remained hidden because affected people lacked the resources or confidence to pursue them.

His argument challenges public bodies to consider whether the apparent increase in complaints represents unnecessary correspondence or a previously underrepresented demand for accountability.

Why the AI Complaint Debate Is More Complicated Than Workload

The disagreement between complaint investigators and AI advocates reflects a wider challenge facing public services as generative AI becomes more accessible.

On one side, organisations must process complaints efficiently. Investigators have limited time and need to identify the central issue, assess evidence and communicate decisions. Extremely long submissions can make this work more demanding, particularly when they repeat the same points or contain legal language that does not directly relate to the dispute.

On the other side, public bodies have a responsibility to provide accessible complaint systems. People should not be disadvantaged because they cannot write a polished letter or understand formal administrative language.

AI can help address this inequality, but it can also introduce new problems when users rely on generated material without checking whether it accurately describes their circumstances.

For example, a chatbot may produce a letter containing legal references that are irrelevant to the complaint, repeat allegations that have already been addressed or present assumptions as established facts. If a user submits the material without reviewing it, the resulting document may make the investigation more complicated rather than clearer.

There is also a risk that repeated use of AI-generated templates could make different complaints sound almost identical, even when the underlying experiences are different. Investigators may then need to spend additional time identifying the individual facts of each case.

These difficulties do not mean that AI-assisted complaints should be rejected. They demonstrate why both complainants and public organisations need to use the technology carefully.

How Public Bodies Can Manage AI-Generated Complaints

Public organisations may need to adapt their complaint-handling systems as AI-assisted correspondence becomes more common. Simply asking people not to use chatbots is unlikely to address the underlying reasons they turn to the technology.

A more practical approach would be to make it easier for people to explain their problems in a clear and structured way, regardless of whether they write the complaint themselves or receive help from AI.

Housing providers and councils could encourage complainants to include the essential details of a dispute, such as:

• The specific problem they want investigated.
• When and where the issue occurred.
• Previous attempts to contact the organisation.
• Any relevant evidence, photographs or correspondence.
• The action or resolution they are seeking.

Such guidance could help reduce unnecessary repetition while preserving the ability of people to submit detailed evidence when their circumstances require it.

Organisations could also use AI to classify incoming correspondence, identify recurring issues, summarise lengthy submissions and direct complaints to the appropriate departments. However, important decisions should remain subject to human review, particularly when complaints involve housing safety, health risks, legal rights or vulnerable individuals.

Automated systems should not determine that a complaint is invalid simply because it is long, uses formal language or appears to have been generated by a chatbot.

Human Judgement Remains Essential in Complaint Investigations

The growing use of AI in complaint writing does not remove the need for human investigators. Many grievances involve circumstances that cannot be understood through document processing alone.

A housing complaint, for example, may involve a tenant’s living conditions, previous conversations with staff, the history of repair requests and the impact of a problem on a household. Investigating such a case requires more than identifying keywords or summarising a letter.

Human staff must assess the available evidence, ask follow-up questions where necessary and understand the outcome the complainant is seeking. They also need to recognise when a seemingly routine complaint may involve a serious safety or welfare issue.

AI can assist with administrative work, but it cannot independently establish whether a landlord has fulfilled its responsibilities or whether a council has provided an adequate service. Those questions require proper investigation and accountable decision-making.

The same principle applies to people submitting complaints. AI can improve the presentation of a grievance, but the individual remains responsible for checking that the information is accurate and reflects what actually happened.

The Future of AI-Assisted Complaints in the UK

The debate over AI-generated complaints is likely to become more significant as chatbots become part of everyday communication. People are increasingly using these tools to write emails, prepare applications, challenge decisions and communicate with institutions.

For public bodies, this means complaint-handling systems may need to accommodate a greater volume of digitally assisted correspondence. The challenge will be to distinguish between legitimate complaints that require attention and repetitive or abusive communication that creates unnecessary pressure on staff.

For tenants and other service users, AI offers a way to overcome barriers that have historically made formal complaints difficult. People who lack confidence in writing or knowledge of administrative procedures can use technology to explain their experiences more clearly.

The three perspectives reveal why neither side of the debate can be ignored. Investigators need manageable workloads and meaningful communication, while complainants need accessible routes to challenge poor services and seek remedies.

Ultimately, the central issue is not whether a complaint was written by a person or an AI chatbot. It is whether the grievance is genuine, whether the facts can be examined and whether the organisation responds appropriately. Public bodies will need to adapt to the changing nature of complaints without allowing technology to become either a barrier to accountability or a source of avoidable administrative pressure.