CompanyGlobalUK10 Sep 2026

Recovering the Costs of AI-Assisted Litigation Support

Applying Established Principles to Modern Practice in Personal Injury and Clinical Negligence.

By Samantha Rowles

Introduction

The effective management of documentary evidence is central to many personal injury and clinical negligence claims.

Solicitors are required to review extensive medical records, organise and analyse that material in order to advise clients, assess breach and causation, instruct experts, prepare evidence and present cases effectively.

The challenge is particularly acute in cases involving significant medical histories or complex issues of causation, where the documentary evidence may extend across many years and involve multiple sources.

Historically, the organisation and review of such material has been undertaken by fee earners, supported where appropriate by specialist litigation support providers offering services such as document management, indexing, pagination and bundle preparation.

Developments in artificial intelligence ("AI") are now introducing new methods of delivering those same litigation support functions.

Platforms such as FinLegal have been developed to assist firms with document-intensive aspects of litigation, including the organisation of medical records, chronology preparation, document categorisation and the creation of searchable electronic bundles.

Importantly, these technologies are not intended to replace legal expertise or professional judgment. Their purpose is to assist solicitors by processing and organising substantial volumes of evidential material more efficiently, reducing the time spent on certain preparatory tasks and allowing greater focus on legal analysis, client advice and case strategy.

AI Throughout the Lifecycle of a Claim

Artificial intelligence has the potential to assist solicitors at different stages of the lifecycle of a claim.

At the earliest stage, AI may assist firms in evaluating whether a prospective claim has sufficient merit to justify further investigation or whether instructions should be accepted. By rapidly reviewing and analysing documentary material, AI may help firms identify potential issues, assess prospects and make more informed decisions about whether a claim should proceed.

The benefits of this type of technology are clear. It may enable firms to assess potential claims more efficiently and avoid significant time being invested in matters that are ultimately unlikely to succeed.

However, the costs analysis in that context raises different considerations.

Work undertaken to determine whether a firm should accept instructions is primarily concerned with case selection and risk assessment. Whether any associated costs are recoverable is therefore a different question from the recoverability of costs incurred in progressing litigation – the issue falls outside the scope of this article but will feature in a future edition.

It considers the position where a claim has been accepted and AI-assisted technology is used to support the conduct of litigation itself for example, by assisting with the organisation of medical records, preparation of chronologies and creation of electronic bundles.

It is those litigation support functions that raise significant questions concerning inter partes costs recovery.

Technology Does Not Alter the Costs Analysis

The history of litigation demonstrates that technological developments do not automatically create new categories of recoverability.

Electronic disclosure platforms provide an obvious example. The introduction of sophisticated document management systems, technology-assisted review and advanced searching methods transformed the way disclosure exercises were conducted.

However, the courts did not approach those costs by asking whether the technology itself was recoverable.

The focus remained upon established questions:

  • Was the expenditure reasonably incurred?
  • Was it proportionate?
  • Was it necessary for the proper conduct of the litigation?

The same approach is likely to apply to AI-assisted litigation support.

Artificial intelligence should not be treated as a special category of expenditure simply because it is innovative in the same way eDisclosure platforms were not.

Instead, the analysis should focus upon the service provided, the purpose for which it was incurred and the established principles governing recovery between the parties.

Applying Established Costs Principles to AI-Assisted Litigation Support

The fact that artificial intelligence has assisted in the delivery of a litigation support service should not, by itself, determine whether the associated costs are recoverable.

The starting point remains the established principles governing costs recovery.

Under CPR Part 44, the Court will allow only those costs which are reasonably incurred and proportionate to the matters in issue. Those principles apply irrespective of whether the work is undertaken manually, through traditional software, by an outsourced provider or with the assistance of artificial intelligence.

The introduction of AI does not create a new category of recoverable expenditure.

Rather, it requires practitioners to apply familiar principles to a developing method of delivering litigation support.

The key questions are therefore likely to be:

  • What service has actually been provided?
  • Was that service reasonably incurred for the purposes of conducting the litigation?
  • Was the cost proportionate?
  • How should the expenditure properly be characterised?
  • Is there a contractual liability to pay for the service?

These are familiar questions within the law of costs and are likely to remain central to any future assessment of AI-assisted litigation support.

Profit Costs, Disbursements and the Indemnity Principle

One of the practical challenges for firms will be determining how AI-assisted litigation support should be presented within a Bill of Costs.

There is unlikely to be a single answer because the correct treatment will depend upon the individual circumstances and the funding arrangements in place.

Where the work forms part of the solicitor's own professional service, recovery may be sought through profit costs via the agency principle. The solicitor remains responsible for reviewing the material, identifying relevant issues and exercising legal judgment.

Where an external provider is engaged to supply a defined litigation support service, consideration may need to be given as to whether the expenditure should be treated as a disbursement.

The correct analysis will depend upon:

  • the contractual arrangements;
  • whether the technology forms part of the firm's general infrastructure or has been engaged as a specific litigation support service for the claim;
  • the basis upon which charges are calculated; and
  • the nature of the work undertaken.

The decision in Crane v Canons Leisure Centre demonstrates that the court may look beyond labels when considering recoverability. Whilst the case concerned external costs drafting services rather than artificial intelligence, the broader principle is relevant: the substance of the service provided is likely to be more important than the terminology used.

The indemnity principle remains central. A receiving party cannot ordinarily recover from a paying party sums which it is not itself liable to pay. Accordingly, firms considering recovery of AI-assisted litigation support should ensure that the contractual arrangements with clients and providers are clear.

The Analogy with Existing Litigation Support Services

Modern disclosure exercises frequently rely upon sophisticated document management systems, searching tools and technology-assisted review processes. These technologies have transformed how parties identify, review and present documentary evidence.

The courts have not treated those services as inherently irrecoverable merely because technology has been involved.

Instead, the focus has remained upon whether the expenditure was reasonable and proportionate in the context of the particular case.

Similarly, specialist providers have long assisted solicitors with functions such as scanning, indexing, pagination, document management and bundle preparation.

The issue has not been whether the work was performed by a solicitor personally.

The issue has been whether the expenditure represented a reasonable cost incurred for the purpose of conducting the litigation.

AI-assisted litigation support should arguably be approached in the same way.

The technology may be different, but the underlying question remains the same.

How should the costs associated with AI-assisted litigation support be treated on assessment?

Although there is currently no reported authority dealing specifically with the recoverability of AI-assisted litigation support, this should be distinguished from the developing judicial and regulatory guidance concerning the responsible use of artificial intelligence within legal practice.

Recent commentary and decisions have focused primarily upon the risks associated with relying upon inaccurate AI-generated material, including the need for proper verification, supervision and professional judgment. Those issues concern the responsible use of AI. They do not determine the separate question of how the costs of AI-assisted litigation support should be treated between the parties.

The question for practitioners is therefore whether existing principles of reasonableness, proportionality, characterisation and the indemnity principle can be applied to this emerging form of litigation support.

Practical Guidance for Firms

Until there is direct judicial guidance, firms adopting AI-assisted litigation support should ensure that their approach is transparent and evidence-based.

Describe the service, not the technology

A Bill of Costs should avoid vague descriptions such as "AI fee" or "technology charge".

Those descriptions do not explain the purpose of the expenditure.

Instead, descriptions should identify the litigation function performed.

For example:

  • AI-assisted organisation and categorisation of medical records;
  • preparation of a chronology from medical documentation;
  • creation of a searchable electronic bundle;
  • document management support for expert review; or
  • AI-assisted disclosure preparation.

Retain supporting documentation

Firms should retain evidence explaining:

  • what service was provided;
  • how charges were calculated;
  • why the service was required;
  • who was responsible for payment; and
  • how the solicitor reviewed and utilised the output.

A clear evidential trail will assist in demonstrating that the expenditure was incurred for the purpose of conducting litigation rather than being a general business cost.

Ensure appropriate supervision

AI-generated material should be reviewed and verified before it is relied upon.

This is not simply a regulatory requirement. It is also important when explaining why the expenditure represents a legitimate litigation support service.

Looking Ahead

AI-assisted litigation support remains a developing area.

There is currently no reported authority determining precisely how these costs should be presented or assessed. However, the absence of direct authority should not prevent practitioners from applying established costs principles.

The introduction of new technology has always required the profession and the courts to consider how existing principles apply to evolving methods of conducting litigation.

AI-assisted litigation support is likely to follow the same path.

The important questions will remain:

  • What service was provided?
  • Was it reasonably incurred?
  • Was it proportionate?
  • How should the cost properly be characterised?

Those are familiar questions within the law of costs.

The technology may be new, but the principles are not.


Samantha Rowles, Senior Costs Consultant, Peak Costs

Samantha is a highly experienced costs lawyer with 19 years' experience in costs litigation. She specialises in complex, high-value personal injury, clinical negligence and commercial disputes, managing matters from inception through to detailed assessment. Samantha has extensive experience acting for both receiving and paying parties, giving her a well-rounded, strategic and commercially focused approach to costs litigation.


This article is general information, not legal or costs advice. Recoverability turns on the facts of the individual case, and nothing here should be relied on in place of advice on your own matter. For costs advice, drafting or budgeting, speak to Peak Costs at peakcosts.com.


Start your FinLegal trial with £3000 of credit at finlegal.ai/get-started

Published 10 September 2026 by Samantha Rowles

More reading

More from the blog.

View all posts →

Try the platform.

See how FinLegal Medical AI helps your team run claims at scale — with £3,000 in free credit and no subscription.

Rejoining the server...

Rejoin failed... trying again in seconds.

Failed to rejoin.
Please retry or reload the page.

The session has been paused by the server.

Failed to resume the session.
Please retry or reload the page.