The use of Artificial Intelligence (or "AI") is on the rise. AI is being used by professionals across various industry sectors.
Legal professionals are making use of AI tools such as Microsoft Copilot and Legora. This is unsurprising given the many uses AI has including:
- Assisting with the drafting of documents (such as contracts and pleadings)
- Analysing and summarising documents
- Clarifying terminology
- Undertaking due diligence exercises
It is inevitable that litigants in person ("LIPs") – individuals who represent themselves in legal proceedings without enlisting the assistance of solicitors and barristers – also turn to AI.
Use of AI by LIPs in litigation
From the pre-action stage through to trial, AI is being deployed by LIPs to assist with tasks that would traditionally fall within the remit of qualified legal professionals.
At the outset of proceedings, LIPs may use AI to research the merits of a potential claim or defence, identify relevant legislation and case law, and draft pre-action correspondence.
Such (such as ChatGPT) can appear as more cost-effective alternative to instructing legal professionals to undertake the same task.
A drawback of using AI to draft pre-action correspondence is that the output produced by AI (whilst, at a surface level, appearing superficially impressive) may lack the concision and precision of effective pre-action correspondence. Use of AI may result in protracted communications between the parties, with the focus shifted to inconsequential or trivial issues, while the core issues in dispute forming the crux of the matter remain unsolved. Given the purpose of pre-action correspondence is to streamline disputes (and minimise costs), use of AI may be counterproductive and result in increasing costs for parties, instead of seeking to narrow issues in dispute. In addition, AI may not be able to grasp all the requirements that need to be covered under a pre-action correspondence.
On 30 June 2026, the Civil Justice Council (CJC) published an interim update detailing the initial findings from its public consultation on the Use of AI for Preparing Court Documents. The consultation, overseen by a working group chaired by Lord Justice Birss, observed that "AI chatbots are now being used by unrepresented litigants. They may be the only source of advice or assistance some litigants receive. Litigants rarely have the skills independently to verify legal information provided by AI chatbots and may not be aware that they are prone to error. If it appears an AI chatbot may have been used to prepare submissions or other documents, it is appropriate to inquire about this, ask what checks for accuracy have been undertaken (if any), and inform the litigant that they are responsible for what they put to the court/tribunal." [1]
The CJC identified areas in which AI can be used by LIPs during legal proceedings: [2]
- Statements of case: AI may be used in connection with the preparation of statements of case to summarise background material, to put it into chronological order, to produce appendices and so forth.
- Disclosure: The Technology & Construction Court was the first court to develop an e-Disclosure protocol. In Pyrrho Investments v MWB [2016] EWHC 256, the court, in approving the use of "predictive coding" in an electronic disclosure process, held (inter-alia) that the parties had agreed on the use of the software and how to use it and there were no factors of any weight pointing away from approving the process. Using software would be appropriate in the instant case and would promote the overriding objective. Since this decision, it has become possible to carry out context searches and to examine documents in efficient and accurate ways, albeit the scope and quality of the search is something that the court will need to grapple with in connection with requests by parties to use increasingly innovative tools to facilitate disclosure.
Other types of drafting should use AI with caution. For example, with regard to witness statements, paragraph 32.1.1.2 of the White Book 2025 provides: “There is currently no specific guidance as to whether generative AI can be used in the preparation of witness statements. However, strong emphasis is placed in the rules on the evidence of the witness being as much as possible in their own words and coming from their personal knowledge. This generally appears to preclude any use of generative AI in drafting them. For example, witness statements need to be (if practicable) in the witness’s “own words".
AI - why is it attractive to LIPs?
In "Artificial Intelligence (AI) Guidance for Judicial Office Holders dated 31 October 2025, key members of the Judiciary provided guidance to judicial office holders and also sought to raise awareness that court and tribunal users, including LIPs, might use AI, for example because this may be the only source of advice or assistance some litigants receive. The guidance notes that some of this may well be useful:
- AI tools are capable of summarising large bodies of text.
- AI tools can be used in writing presentations, e.g. to provide suggestions for topics to cover.
- Administrative tasks can be performed by AI, including composing, summarising and prioritising emails, transcribing and summarising meetings, and composing memoranda.
AI - is it all glorious?
Courts have encountered cases in which users (including both legal professionals and LIPs) have submitted AI-generated documents containing fabricated case citations and inaccurate legal propositions. This has prompted a broader discussion as to the extent to which reliance on AI is beneficial, and the circumstances in which such reliance may, in fact, prove detrimental to a party's case. While legal professionals must use any AI subject to their duties to the Court and within their regulatory responsibilities and requirements, there is potentially more risk associated with AI use by an LIP.
One such case concerning an LIP is BEA, R (on the application of) v Staffordshire County Council [2026] EWHC 1190 (Admin). The case concerned a claimant's judicial review challenging Staffordshire County Council's placement of her son with his paternal aunt between August 2025 and February 2026.
On AI-generated submissions, Kimblin J noted his concern that the claimant's written materials may have been AI-generated, after noticing the word "academicity" being referenced in the same way in two separate litigants-in-person cases before him in the same week. The Judge noted: "The basis of the claimant’s written submissions is not known and I am sceptical about it. The court has to have confidence in the research and comprehension of the legal authorities which underpin a submission. My concern that AI has contributed to the written argument without reference to the key authorities may be ill-founded.” [3]
Kimblin J is not the only judge to be sceptical about the use of AI by LIPs. In Zzaman v Revenue and Customs (HICBC) [2025] UKFTT 539, the Tax Tribunal opined that "Litigants using AI tools for legal research would be well advised to check carefully what it produces and any authorities that are referenced. These tools may not have access to the authorities required to produce an accurate answer, may not fully “understand” what is being asked or may miss relevant materials. When this happens, AI tools may produce an answer that seems plausible, but which is not accurate." [4]
It is well known that AI tools have a tendency to produce outputs that align with what the user wishes to hear, rather than necessarily providing an objective and balanced assessment. For example, if a LIP prompts an AI tool by asking whether they have a strong claim for breach of contract, the AI may generate a response that emphasises the strengths of the LIP's position whilst downplaying or omitting significant obstacles, such as limitation issues, evidential deficiencies, or contrary authorities. Similarly, a LIP seeking to resist an application for summary judgment may receive AI-generated submissions that present arguments in favour of the LIP's position without adequately addressing the applicable legal test or the strength of the opposing party's case. AI may also selectively identify case law that appears to support the LIP's desired outcome, without drawing attention to distinguishing facts or subsequent authorities that may undermine the relevance of the cited decisions. The risk is that a LIP, lacking the legal training to critically evaluate the AI's output, may proceed on the basis that their claim is more meritorious than it is, only to be met with adverse costs consequences or the summary disposal of their claim or defence.
In Hancox v Sutherland [2026] EAT 139, a LIP submitted a 300-page skeleton argument created using ChatGPT. The case concerned Mr Hancox, arguing (inter-alia) that he had not been adequately warned that a strike-out application would be decided at a hearing in 2021. Mr Hancox's appeal initially failed at the EAT’s preliminary assessment, but a further preliminary hearing was eventually ordered to decide whether there was an arguable basis for it to proceed.
The judge criticised Mr Hancox for not complying with the Practice Direction of the Employment Appeal Tribunal 2024 and served a skeleton "to obscure rather than to illuminate during preparation for the hearing."[5] Whilst the judge ultimately considered there were reasonable grounds for bringing the appeal (on one ground), the judge's comments on the use of AI is nevertheless useful in ascertaining the court's views as regards AI:
"All litigants and their representatives, including litigants without representation, should be mindful of the risks posed by the use of generative AI when creating documents for submission to the EAT. They must take personal responsibility to ensure, at a minimum, that documents (i) comply with applicable procedural rules, (ii) have been checked as thoroughly as the litigant or representative is reasonably able for accuracy, ensuring that the factual, evidential and legal position is not misstated and the EAT is not misled, and (iii) contain only relevant points, with the focus being on the central or best arguments, presented in an easily comprehensible manner and avoiding undue repetition. It is not acceptable to submit documents created by generative AI without them being checked, or to attempt to place the onus for checking them on the opposing party." [6]
AI tools, which lack the capacity to grasp the nuances and interplay of complex regulatory frameworks may generate results that appear authoritative, but which fail to account for critical procedural requirements or recent developments in the law. The risk is compounded in circumstances where a LIP, lacking the expertise to identify such deficiencies, proceeds to act upon AI-generated guidance without further scrutiny. Practitioners will also be aware that AI has the capacity to hallucinate authority to support a particular position, as explored in our article here.
The use of AI may also give rise to a further, and potentially more serious, consequence: the wasting of the court's time and the imposition of civil restraint orders. Where AI-generated submissions contain fabricated authorities, legally unfounded arguments, or repetitive and unmeritorious applications, the court's resource may be diverted from the efficient administration of justice. A LIP who repeatedly files AI-generated claims or applications that are found to be totally without merit may be subject to a civil restraint order ("CRO") under CPR 3.11 and Practice Direction 3C. A CRO, if issued, restricts a party's ability to issue further claims or make further applications without first obtaining the permission of a judge. CROs may be imposed where the court is satisfied that the litigant has persistently issued claims or made applications which are totally without merit. LIPs should therefore exercise considerable caution before relying upon AI-generated content to initiate or sustain proceedings and should be aware that an uncritical use of AI may not only undermine their case but also jeopardise their future ability to access the courts.
Comment
There is no doubt that AI has significant uses for the professional sectors and individuals alike. The capacity of AI to process vast amount of information at speed is a welcome development in the accessibility of analysing complex information and provision of assistance that could otherwise be unavailable to an LIP. For legal professionals, where suitable to a matter, AI offers the prospect of enhanced efficiency and the ability to allocate resources more strategically.
The Lady Chief Justice of England and Wales, Baroness Carr of Walton-on-the-Hill has commented that, as per the guidance discussed above, "we as judges are trained to look out for fake hallucinations, fake cases and the like, many judges are saying to be that they find AI-assisted submissions from litigants in person more helpful, easier to digest than submissions that are drafted without the support of AI". The Lady Chief Justice opined that "AI, we are seeing, can really help litigants in person have access to justice." [7]
However, AI, if not used properly and with care can pose risk, particularly where litigation is concerned including creating information which is inaccurate, misleading and simply false. Whilst the courts have some sympathy for LIPs as regards the use of AI, lack of legal knowledge will unlikely absolve LIPs for reckless and improper use of AI.
The Courts and Tribunals Judiciary has provided the following guidance for responsible use of AI in Courts and Tribunals [8]:
- Understanding AI and its applications: Before using any AI tools, it is important to grasp an understanding of the AI's capabilities and potential limitations. Whilst AI tools may be useful to find material one would recognise as correct ;it is a poor way of conducting research to find new information that cannot be verified. The quality of any answers received will depend on how one engages with the relevant AI tool, including the nature of the prompts entered, and the quality of the underlying datasets These may include misinformation (whether deliberate or otherwise), selective data, or data that is not up to date. Even with the best prompts, the information provided may be inaccurate, incomplete, misleading, or biased. It must be borne in mind that “wrong” answers are not infrequent.
- AI tools do not uphold confidentiality and privacy. The guidance suggests that confidential information should not be inputted in AI chatbots.
- The accuracy of any information provided by an AI tool must be checked before it is used or relied upon, as information provided by AI tools may be inaccurate, incomplete, misleading or out of date.
- AI tools may have a tendency to generate responses based on the dataset they are trained upon. Information generated by AI will inevitably reflect errors and biases in its training data, perhaps mitigated by any alignment strategies that may operate.
The Tax Tribunal's comments in Zzaman v Revenue and Customs provides a useful insight into how to mitigate risks when using AI. The Tribunal noted that "dangers can be reduced by the use of clear prompts, asking the tool to cite specific paragraphs of authorities (so that it is easy to check if the paragraphs support the argument advanced), checking to see the tool has access to live internet data, asking the tool not to provide an answer if it is not sure and asking the tool for information on the shortcomings of the case being advanced. Otherwise there is a significant danger that the use of an AI tool may lead to material being put before the court that serves no one well, since it raises the expectations of litigants and wastes the court’s time and that of opposing parties." [9]
Drawing upon our experience, we offer the following practical tips for dealing with AI-generated pre-action correspondence:
- Identifying the core issues: AI-generated letters often bury the substantive points within layers of unnecessary detail. It is advisable to distil the correspondence to its essential elements and consider responding only to those points which are material to the dispute.
- Keeping responses concise and focused: resist the urge to address every point raised in a lengthy AI-generated letter. A disciplined and succinct response is more likely to advance the pre-action process efficiently and demonstrate to the court, if proceedings are subsequently issued, that reasonable steps were taken to resolve the dispute proportionately.
- Where appropriate, flagging concerns regarding AI-generated content: if it is apparent that correspondence has been generated by AI and contains inaccuracies, fabricated references, or irrelevant material, it may be appropriate to draw this to the attention of the opposing party at an early stage. Doing so may encourage a more constructive dialogue and reduce the risk of further AI-generated content being relied upon.
- Maintaining focus on pre-action protocol compliance: regardless of the volume or quality of AI-generated correspondence received, parties should ensure that their own pre-action conduct remains compliant with the relevant pre-action protocol. The court will assess the reasonableness of the parties' conduct by reference to the protocol requirements, not the volume of correspondence exchanged.
For LIPs, practical tips might include:
- LIPs should, at a minimum, independently verify any case law citations or legal propositions generated by AI against authoritative legal databases to ensure their accuracy and continued validity.
- Where AI is used to draft legal submissions, LIPs ought to critically review the output rather than submitting AI-generated content verbatim, as AI tools may produce text that appears superficially persuasive, but which lacks substantive legal foundation. In particular, LIPs should be alert to the tendency of AI tools to adopt a verbose approach, generating submissions that raise every conceivable argument irrespective of merit, rather than advancing focused and proportionate submissions directed to the issues that are genuinely in dispute.
- LIPs should be mindful that AI tools are not a substitute for qualified legal advice, particularly in respect of matters involving complex legal principles and procedural requirements. Where possible, LIPs should seek to obtain legal advice. If — whether through pro bono services, law clinics, or other accessible avenues if unable to pay to access — to complement any AI-assisted research.
Whilst AI can be a blessing, can no doubt make real practical difference to litigants, and has the potential to lower the cost of obtaining legal advice, given the warnings given by the judiciary discussed above, LIPs should ensure it does not turn into a curse by ensuring they adopt good practices in connection with the use of AI in any legal disputes or proceedings.
[1] Referred to in the Civil Justice Council: 'Use of AI for Preparing Court Documents -Interim Report and Consultation'.
[2] Civil Justice Council: 'Use of AI for Preparing Court Documents -Interim Report and Consultation'
[3] BEA, R (on the application of) v Staffordshire County Council [2026] EWHC 1190 [paragraph 38]
[4] Zzaman v Revenue and Customs (HICBC) [2025] UKFTT 539 [paragraph 29]
[5] Hancox v Sutherland [paragraph 2]
[6] Hancox v Sutherland [paragraph 4]
[7] Lady chief justice commends AI usage by litigants in person, the Law Society Gazette
[8] Courts and Tribunals Judiciary, Artificial Intelligence (AI) Guidance for Judicial Office Holders, 31 October 2025
[9] Zzaman v Revenue and Customs [paragraph 29]
