Generative AI in Complaints and Requests: Opportunities, Risks and Implications for Public Bodies

George McLellan, Fred Groves & Christopher Watkins discuss the impact of generative AI on the publics' interactions with public bodies.

Introduction

Generative AI can offer a host of benefits to members of the public when interacting with public bodies, particularly to individuals who may previously have faced difficulties in composing formal correspondence, such as a complaint or request for information. However, the increasing use of generative AI is placing strain on public bodies, who are under legal duties to properly consider, verify and meaningfully respond to complaints and requests. The challenge for public bodies is in maintaining fairness, consistency and transparency when handling increasing volumes of AI-generated submissions.

Access to justice, information and redress

Used effectively and responsibly, generative AI can improve access to justice. By way of example, a recent Civil Justice Council interim report acknowledged the potential benefits of generative AI technology to those without legal representation in court proceedings (litigants in person) in the preparation of court documents, such as particulars of claim and witness statements.

Until recently, the requirement for a written submission in support of a complaint or request has proved a barrier to some. Participating in formal complaints and request processes can be daunting, and unfamiliarity with the procedural requirements can deter potential complainants or requesters. A person may have a legitimate grievance but find it hard to articulate and present their position in the form of a written complaint. Many do not have the time to prepare a detailed submission, or lack the expertise required in some cases (for example, experience in the relevant field of law or regulation).

Generative AI tools can help users to strengthen their written arguments, to introduce legal analysis in support of their objectives and to better understand and navigate the process of interacting with a public body. From a public interest perspective, this can represent a genuine benefit to both parties. Well prepared complaints (including those drafted with the use of generative AI) can help public bodies to better identify systemic issues, assist regulators in spotting patterns of poor practice and support ombudsmen in resolving disputes more efficiently. The depth and comprehensiveness of responses composed using generative AI can in some cases aid the identification of the relevant issues and remedies. In other cases, the use of generative AI tools can frustrate the complaint and request processes.

Emerging risks

Recent UK government data show that the number of Freedom of Information (FOI) requests continues to increase across public bodies. The total number of FOI requests rose from just under 15,000 per quarter in 2022 to over 24,000 per quarter in 2025. Whereas the majority of FOI requests made to UK government bodies have been responded to within the statutory timeframe, the data nonetheless highlights the scale of material that must be processed, assessed and addressed by public bodies, many of whom face budgetary constraints. We note that the recent increase in FOI requests coincides with the popular adoption of generative AI tools in recent years, following the release of large language models such as Chat GPT in November 2022. Whilst the proportion of FOI requests created using AI tools is uncertain, we suspect that the rise in the number of FOI requests may be linked. Text can now be generated instantaneously and free of charge. As a result, some requesters may now be more inclined to submit multiple requests to several public bodies simultaneously. This in turn increases the pressures faced by the responding public bodies.

Generative AI is contributing to a marked increase in the volume, length and complexity of complaints received by public bodies. Schools are reporting an ‘exponential’ increase in AI-generated parental complaints that, when compared with non-AI complaints, tend towards being more voluminous, more antagonistic in tone and more likely to threaten escalation or litigation. This places additional pressure on teachers and administrators, who may struggle to resolve those complaints informally and amicably as a result. Regulators and ombudsmen describe receiving an increase of AI-generated submissions that resemble lengthy legal documents, even in cases where the issues are relatively straightforward and do not warrant that level of complexity and formality.

A risk of this trend is that genuine issues become obscured or buried beneath lengthy and irrelevant material, meaning that public bodies are required to allocate significantly more time, resources and cost in dealing with these complaints and requests. The volume, complexity and risk of error in some AI-generated content (as discussed below) may also undermine the complainant’s objectives.

Hallucination, distortion and inaccuracy

A further difficulty lies in the reliability of AI‑generated content. Generative AI is capable of producing ‘hallucinations’ in its output. An AI‑generated complaint or request may:

• Include erroneous of fictitious references to case law;
• Cite the wrong statutory regime;
• Misstate the scope of a public authority’s legal duties;
• Attribute decisions or statements that were never made; or
• Conflate different incidents into a single narrative.

The responsible use of generative AI technology involves the careful review and verification of AI outputs. However, recent case law such as Ayinde v London Borough of Haringey [2025] EWHC 1040 (Admin) highlights the risk of failing to check AI outputs for accuracy and relevance. In Ayinde, the court found that a party had cited non-existent or materially misstated authorities generated using AI tools in its statements of case.

Because of the apparent sophistication of generative AI outputs, some users may be lulled into considering proper checking and verification to be altogether unnecessary. AI-generated content may seem at first glance to be coherent and properly sourced but can fail to stand up to proper scrutiny. The inherent risk of inaccuracy and hallucinations (as discussed below) in AI-generated outputs increases the existing demands placed upon public bodies, who are required to exercise due diligence and caution when dealing with complaints and requests of all types.

AI responding to AI

In response to the growth in AI‑generated complaints, some public bodies will consider the use of AI tools to manage the complaint or request process itself. Schools Week reports that around 40 schools are using an AI‑assisted platform to log, categorise and manage complaints, with internal AI tools used to assist in drafting responses. In the regulatory sphere, the Local Government and Social Care Ombudsman has launched an AI assistant to “[…] assist prospective complainants on how to make a complaint and what the Ombudsman can and cannot look at’. These developments reflect a broader move towards AI‑supported complaints handling across both the public and private sectors.

These tools may offer a more consistent, efficient and cost-effective approach to complaint handing. AI can be adept at categorising, analysing and preparing responses to complaints and submissions, provided there is sufficient human oversight. However, this approach is not without risk. AI tools may struggle to interpret nuance or fail to identify safeguarding or equality issues. For schools, regulators and ombudsmen – where complaints frequently engage statutory duties, legal rights or reputational interests – decisions must remain clearly attributable to, and defensible by, human decision‑makers. Unless strong governance and oversight mechanisms and established, an AI‑assisted response to an AI‑assisted complaint risks compounding errors or undermining the robustness and transparency of the process.

Public law duties in complaints handling

The growth in AI‑assisted submissions does not alter the legal framework within which public bodies must operate. Public bodies remain bound to exercise their powers lawfully, to follow their published procedures, and to reach decisions that are lawful, procedurally fair, rational and grounded in an accurate understanding of the facts.

We consider that the use of generative AI increases the likelihood that irrelevant considerations, factual errors or misunderstood legal duties are taken into account when handling complaints and requests. Decisions made on this basis are vulnerable to challenge on grounds of procedural unfairness or irrationality.

Published complaints procedures may give rise to legitimate expectations as to how complaints will be handled. Inconsistent treatment of complaints where, for example, there has been an undisclosed, inconsistent or irrational application of AI tools could undermine fairness, particularly where the resulting decisions affect reputation, safeguarding or statutory entitlements.

Proportionality is another consideration for public authorities when assessing how to respond to high volumes of complaints and requests. Introducing targeted policies to mitigate the risks of AI generated content may be considered proportionate in circumstances where the absence of those policies would be detrimental to local authority resourcing. For example, organisations may have considered implementing word limits, AI identification tools or AI use declarations to better manage their complaint or request processes. However, there is no ‘one size fits all’ solution, and the right balance must be struck.

How Sharpe Pritchard can help

The recent proliferation of generative AI is reshaping the way individuals and organisations engage with public bodies, including schools, regulators and local authorities. Many public bodies face constraints on the time and resources available for handling complaints and requests, and will be keen to implement changes to maximise the efficiency of these processes. Those changes should be driven by clear and transparent policy. Public bodies should remain mindful of their legal and ethical duties, and may benefit from legal advice and support when considering how best to adapt to the rapid pace of technological developments in this field.

If you would like further advice and assistance in relation to any of the issues raised in this article, please contact us today by telephone or email This email address is being protected from spambots. You need JavaScript enabled to view it. and This email address is being protected from spambots. You need JavaScript enabled to view it..

George McLellan is a Partner and Fred Groves & Christopher Watkins are Associates at Sharpe Pritchard LLP.

---

For further insight and resources on local government legal issues from Sharpe Pritchard, please visit the SharpeEdge page by clicking on the banner below.

 

 

Visit Sharpe Pritchard's new Building Safety Hub, focusing on The Building Safety Act 2022 and its wide-ranging impact.

This article is for general awareness only and does not constitute legal or professional advice. The law may have changed since this page was first published. If you would like further advice and assistance in relation to any issue raised in this article, please contact us by telephone or email This email address is being protected from spambots. You need JavaScript enabled to view it..