Published on May 4, 2023
ChatGPT and the Future of Dispute Resolution: The Benefits and Risks of Using AI

Artificial intelligence (AI) is transforming industries at a rapid pace, and the legal sector is no exception. The use of AI in law firms has increased in recent years as firms seek ways to reduce costs while improving efficiency and maintaining quality of service.

Dispute resolution can be time-consuming and expensive, particularly in complex cases. As a result, lawyers and other legal professionals are constantly searching for ways to streamline the dispute resolution process and reduce costs for their clients. One potential solution is to use large language models (LLM), for example, chatbots like ChatGPT, an AI-powered language model that can process natural language input and generate human-like responses. It is based on a deep learning neural network trained on a vast amount of data from the internet. The model can generate responses to various topics, including legal issues.

By automating some aspects of the dispute resolution process and providing objective analysis, ChatGPT can potentially reduce litigation costs in the future. However, there are also challenges and limitations to using ChatGPT in this context.

Use of ChatGPT in Litigation and Arbitration

There are many potential uses of LLM in the legal profession, from answering basic questions to providing legal advice and drafting documents. However, the most valuable uses of ChatGPT in dispute resolution are to assist lawyers with (a) drafting; (b) document review; (c) predicting case outcomes to inform strategy; and (d) settlement negotiations.


As an LLM, ChatGPT can be used to assist lawyers to draft correspondence, legal briefs, applications and pleadings.

Legal writing can be time-consuming and requires high precision and attention to detail. ChatGPT can help automate some aspects of legal writing, such as drafting standard pleadings, by suggesting language based on the facts of the case and the relevant legal authorities. This can save lawyers time and reduce the risk of errors in their pleadings.

Recent research suggests that in simple cases and with appropriate instructions, the quality of legal drafting by ChatGPT can be as of a first-year lawyer. As an experiment, one Australian lawyer asked ChatGPT to draft a statement of claim based on the 1992 Mabo case and was impressed. Another study found that ChatGPT has proficient legal drafting abilities for different kinds of documents such as demand letters, without-prejudice letters, and pleadings. ChatGPT could elaborate and improve the contents of these documents based on the basic facts fed into the system and showed the capacity to comprehend simple facts and express the legal justification of the claim.

Human supervision is essential when ChatGPT is used for legal drafting for several reasons.

First, ChatGPT is not a substitute for human judgment and legal expertise. A human lawyer can consider nuances, context, and other factors that a machine might not be able to comprehend.

Secondly, such drafting may contain factual and legal errors. For example, on one occasion, ChatGPT made up a case which did not exist (including a case name, citation and legal principal) to fit the narrative of the drafting task it was asked to perform. Thus, human supervision is essential to ensure the information is accurate and correct any errors the system might make.

Finally, legal drafting requires more than just producing a document with correct syntax and grammar. A human lawyer can understand the nuances and the overall tone of the document and tailor it to the specific audience, taking into account the cultural and social context of the case.

Therefore, the work produced by ChatGPT may be useful in some circumstances to prepare a first draft to be checked and improved by a qualified lawyer, but it should only be deployed with rigorous checks by solicitors familiar with the case and law to ensure that the final product is of high quality and fit for purpose. The extent of the checks that might be required on a given case, may limit the use of the LLM in the first place.

Disclosure and Document Review

ChatGPT can be used in litigation by assisting with document review. Document review is a crucial and often time-consuming aspect of litigation. Until relatively recently, document review was typically only done manually by a team of lawyers, which was costly and prone to errors and inconsistencies.

Courts in England and Wales recognise and approve the use of Technology Assisted Review (TAR), which includes all forms of document review that may be undertaken or assisted by the use of technology, including but not limited to predictive coding and computer assisted review (PD 57AD, Appendix 1, ¶1.15). The judges expect parties to agree on the use of such technology but are prepared to make rulings if they cannot agree.

TAR is increasingly used in large litigations, and many document review platforms offer it as part of the standard service.

LLMs can be used to improve the quality of the AI-assisted review. For example, it can help lawyers automate document review aspects, such as identifying relevant documents, categorising them based on their relevance to the case, and identifying privileged information.

ChatGPT’s ability to process natural language input and generate human-like responses can be used to summarise the main points of a long document in a concise and coherent format. However, the accuracy of the summarisation would depend on the quality of the training data used to develop the model and the complexity of the document. Therefore, it would be necessary to carefully evaluate the accuracy and reliability of ChatGPT’s summarisation capabilities before using it for document review in a legal context.

Using LLM can significantly reduce the time and costs associated with document review, allowing lawyers to focus on more substantive aspects of the case. However, AI can only be used with initial setup, training, ongoing supervision, and quality checks by humans. AI is not a substitute for human expertise and judgement; human oversight is still needed to ensure accuracy and fairness in the review process.

Predicting Case Outcome

Another potential application of ChatGPT in litigation and arbitration is predicting the outcomes of cases to assist in informing strategic direction. Of course, no one can predict the outcome of particular litigation or arbitration with 100% certainty, but the quality and quantity of the data matter.

ChatGPT can analyse large amounts of legal data, including case law and case summaries, to predict how a particular case will likely be decided. This can help lawyers and clients make more informed decisions about settling a case or taking it to trial.

However, it is essential to note that ChatGPT is not infallible, and there is always a degree of uncertainty involved in legal decision-making. Additionally, the accuracy of ChatGPT’s predictions will depend on the quality of the training data used to develop the model and the complexity of the legal concepts involved in the case. The current version of ChatGPT has limited knowledge of the events (including cases) that took place after 2021, and it might be missing crucial legal developments.

Therefore, while ChatGPT has the potential to be a valuable tool for predicting case outcomes, it should be used in conjunction with human judgment and expertise. Ultimately, legal decision-making involves a range of factors beyond just predicting the outcome of a case, including strategic and ethical considerations and client goals.

Settlement Negotiations

Another potential use of ChatGPT in litigation is in assisting with settlement negotiations. Settlement negotiations can be complex and emotionally charged, with both parties trying to negotiate the best possible outcome for themselves.

ChatGPT can assist by analysing large amounts of data related to similar cases and providing insights into potential settlement amounts and negotiation strategies.

It can also assist with identifying potential areas of compromise and suggest creative solutions to help both parties reach a settlement agreement.

Additionally, ChatGPT can be used to simulate various negotiation scenarios and help lawyers develop and test their negotiation strategies before engaging in actual negotiations.

By leveraging the power of natural language processing, ChatGPT can help lawyers make more informed decisions during the settlement negotiation process, potentially leading to faster and more favourable outcomes for their clients.

While ChatGPT can provide valuable insights and suggestions for settlement negotiations, it should not be solely relied upon, and human expertise and judgment are always crucial in the negotiation process. Therefore, ChatGPT’s suggestions should be considered as a tool to supplement and inform decision-making rather than as a substitute for it.

Cost Savings for Clients

One of the most significant advantages of using ChatGPT is cost savings for clients. With ChatGPT, law firms can potentially reduce the time spent on various tasks essential in dispute resolution.

Conducting legal research is an indispensable part of any litigation process. However, it can be time-consuming and expensive. With ChatGPT, lawyers can quickly find relevant case law, statutes, and regulations, reducing research time and legal fees.  Again however, the quality of the output is only as good as the scope of the data, which is relied on, and human resource remains vital in verifying the answer identified.

Legal drafting can be time-consuming and requires attention to detail. ChatGPT can help automate some aspects of legal writing, such as drafting standard pleadings, letters, and other legal documents. By suggesting language based on the facts of the case and legal authorities, ChatGPT can save lawyers time and reduce the risk of errors, ultimately saving clients’ money in legal fees.

Disclosure is often one of the most expensive parts of the litigation process. ChatGPT can assist with e-discovery, automating document review and analysis, and reducing the time and cost involved in the process.

By identifying potential areas of compromise and suggesting creative solutions, ChatGPT can help both parties reach a settlement agreement faster and more efficiently.

While there are some challenges and limitations to using ChatGPT in litigation, its potential benefits make it a tool that should be thoughtfully considered by lawyers and legal professionals looking to reduce litigation costs.

Risks Associated with the Use of ChatGPT

Despite the benefits of using ChatGPT, there are also risks associated with its use, including potential mistakes and bias, protection of confidentiality and copyright breaches.

One obvious risk, identified above, is the potential for errors in the data input. ChatGPT relies on accurate data to generate accurate results, and if the data input is incorrect or incomplete, the results generated by ChatGPT may be flawed.

Another risk is the potential for bias. ChatGPT is trained on large datasets, and if those datasets are biased, the results generated by ChatGPT may also be biased. This could lead to incorrect legal advice or even legal disputes.

Privacy and confidentiality are significant concerns when it comes to using ChatGPT by litigation lawyers. When using ChatGPT for legal drafting or research, lawyers must be cautious when entering client-specific information into the system. In addition, using pseudonyms or fictitious data when feeding ChatGPT with confidential information is essential to prevent data leaks or breaches. It is also unclear whether privilege may be waived by providing information to a chatbot, and this question is likely to come before the courts in the future.

There are potential copyright concerns when using ChatGPT for legal drafting, particularly if the model is trained on copyrighted legal documents or materials. In some jurisdictions, there may be limitations on the extent to which copyrighted works can be used for commercial purposes, including using such works to train machine learning models like ChatGPT. Lawyers should be cautious when using ChatGPT to generate legal documents to ensure that the resulting work does not infringe on the copyright of others.


The legal profession has been exploring ways to use artificial intelligence to streamline processes and reduce costs and no doubt will always continue to do so. Using an LLM, such as ChatGPT, to assist with various aspects of dispute resolution, including drafting, document review, predicting case outcomes, and settlement negotiations offers a potential solution to streamline aspects of the legal function and to reduce costs. ChatGPT has shown proficiency in drafting various legal documents and assisting with document review.

However, it is not a substitute for human judgment and legal expertise and the more complex the litigation at issue, the less likely it is to be effective. Human supervision and active engagement remain essential to ensure that the work produced by ChatGPT is of high quality and fit for purpose.

P.S. This article was created with the assistance of ChatGPT. While ChatGPT provided suggestions and generated some content for this article, the final product was reviewed and edited by a human writer to ensure accuracy and clarity.

Jun 13, 2024
Fair winds favour plaintiffs in wasted expenditure cases
A recent judgment of the High Court of Australia has clarified aspects of contract law and the availability of damages...
Jun 3, 2024
LIDW 2024: The Impact of Satoshi Nakamoto on English law
Over the past few years, the novel Blockchain technology has had a significant disruptive impact on many areas, not least...
May 20, 2024
LIDW 2024: Clash of Jurisdictions – era of sanctions, anti-suit and anti-anti-suit injunctions?
Enyo Law is co-hosting an event at London International Disputes Week 2024 alongside One Essex Court, ALRUD law firm, and...
May 16, 2024
MUR Shipping: Force Majeure and Reasonable Endeavours – Contractual Certainty and how to Foresee the Unforeseeable
This week, the Supreme Court issued its decision in the highly anticipated RTI Ltd (Respondent) v MUR Shipping BV (Appellant)...