The use of Generative Artificial Intelligence (‘Gen AI’) has become almost omnipresent in our world. Well known AI tools such as ChatGPT, which was only released in 2022, have rapidly evolved and have the ability to do things we would not have imagined just 10 years ago.
The use of Gen AI can work as a transformative tool for the legal system, making processes cheaper, quicker and more accessible, but it must be used with caution. The regulation of technological processes, in particular, AI, presents a profound challenge for the legal system. As put by Dame Victoria Sharp, President of the King’s Bench Division of the High Court of Justice, ‘Artificial Intelligence is a tool that carries with it risks as well as opportunities’.
We can see that other technological advancements that have been adopted by the legal system have vastly improved costs, speed and access to justice. Simple processes such as video calls via Zoom or similar platforms for remote hearings and signing of documents are an example of this. AI can do the same, and more, but not quite yet.
What is Generative AI?
Generative AI is a form of artificial intelligence that is capable of creating new content, including text, images or sounds, based on patterns and data acquired from a body of training material. Generic Gen AI programs include ChatGPT, Co-Pilot, Grok, and Claude. More tailored legal AI programs include Lexis+ AI (Protégé) and Westlaw Precision.
What is and is not currently allowed?
Gen AI is already so widely used in the rest of society, it would be a failure of the legal system to disregard its use altogether. The Supreme Court of NSW has taken this into account in guiding how Gen AI may and may not be used in Supreme Court Practice Note SC GEN 23.
According to this practice note, acceptable uses of AI include:
Generation of chronologies, indexes and witness lists;
Preparation of briefs of draft Crown Case Statements;
Summarising or review of documents and transcripts; and
Preparation of written submissions or summaries of argument (provided that the author verify, without using AI, that the contents and all references exist, are accurate and are relevant to proceedings)
Generative AI cannot be used for generating the contents of affidavits, witness statements, character references or other material that is intended to reflect the deponent or witness’ evidence and/or opinion, or preparation of expert reports.
It is vital that solicitors follow these guidelines to ensure the effective use of AI.
How does this help both lawyers and clients?
When AI is used correctly, it has the ability to assist lawyers and clients in many ways, facilitating quicker, cheaper and more just outcomes.
From the solicitor’s standpoint, AI programs can make processes quicker and easier. AI can significantly reduce the time involved in certain tasks including sorting through large volumes of discovery, drafting documents and creating chronologies. This reduction in time also assists the client as it leads to lesser costs and disbursements.
Additionally, AI is a great starting point, emphasis on ‘starting point’, for legal research and summarising cases, especially when more specialised legal AI is used. Programs such as Lexis+ Protégé, are great tools for solicitors and law students, as long as the information is independently verified.
For individuals who need to interact with the legal system who lack legal knowledge, AI can be useful to improve access to justice. AI can provide simplified overviews of the law, a starting point for prospective clients, and may also be used by unrepresented litigants in building their case. However, it is important to note the shortcomings of AI in these areas as information is not always accurate and reliable (see discussion of May v Costaras later). Access to justice is additionally improved by making engaging solicitors cheaper, as discussed above.
Concerns with Gen AI, what do I need to be aware of?
Most people would be aware that Gen AI can quite often have ‘hallucinations’ or include irrelevant information. Gen AI may also have inherent bias in its code that will impact its output.
The drawbacks of this can be seen in May v Costaras [2025] NSWCA 178 where an unrepresented litigant used AI in the preparation of their oral submissions. While displaying the possibility of AI usage by unrepresented litigants, hopefully improving access to justice, the case warns of the possibility of ‘unintelligible’ and ‘irrelevant’ outputs from AI. Chief Justice Bell explains that this case ‘illustrates the need for judicial vigilance in the use of Generative AI, especially but not only by unrepresented litigants’.
Furthering this point, it would be remiss not to include the example of the Victorian lawyer who was the first in Australia to face professional sanctions for using AI in a court case. The lawyer had used AI to generate a case list but failed to verify its contents, leading to false citations being submitted to the court. He was subsequently stripped of his ability to practise as a principal lawyer, among other conditions. This provides a harsh reminder to practicing solicitors on the penalties of incorrect AI usage.
The use of Gen AI also poses a threat to confidentiality. Information that is fed into AI tools may be used to train the program, potentially leading to confidential information being available to others.
In addition to the obvious drawbacks of AI content reliability, a concern for its future use is the removal of the human aspect of the legal system. We do not know how AI may be used in the future, some even suggest that it can be used to partially replace judges in certain circumstances. This is a scary concept and one that should not be taken lightly, especially with regard to the public trust and perception of the courts.
What might AI use look like in the future?
It’s hard to predict exactly how AI will be used in future practice. It can extend from its current research and admin uses, spreading into every aspect of the law. For now, incremental acceptance of AI usage in law is the way forward to ensure it is regulated properly and used effectively. It is imperative that lawyers keep their professional and ethical obligations front of mind in this process.
Ultimately, the use of AI by solicitors and clients alike, is undeniably useful but must be used with caution and in line with current regulations.
This article is general in nature and is not intended to be tailored legal advice.
