Is Artificial Intelligence the New Junior Lawyer?

Introduction

Artificial intelligence (AI) is set to revolutionise and streamline the way we practice law. Complete the work of a junior associate in minutes, whispers Thomson Reuters’ Co-Counsel. Get the legal summary you need in seconds without clicking into a single result, murmurs LexisNexis’ Lexis+AI. The allure of time and cost saving promised by AI has piqued the attention of law firms and their clients around the world. 

Some law firms have been quick to embrace AI, incorporating a mix of general-purpose AI (GPAI) tools like OpenAI’s ChatGPT and Microsoft’s Copilot, and specialised legal AI tools like CoCounsel and Lexis+AI, into daily business processes. Others remain cautious, finding these tools not yet up to scratch¹ or preferring to wait until concerns about bias, reliability and security (amongst other things) have been resolved.² 

Notwithstanding the variety of views on AI, the consensus is clear: AI is here to stay, and it has the potential to reshape the junior lawyer’s role. AI presents an opportunity for all lawyers – junior and senior – to come together and explore different (and perhaps, better) ways to work, and it is up to industry leaders to facilitate the process. 

AI in the Legal Profession

In July 2023, joint research by Microsoft and the Tech Council of Australia found that 10% of a lawyer’s work could be automated and 32% could be augmented by AI in Australia.³ In the United States, Goldman Sachs estimated in March 2023 that 44% of legal work could be undertaken by AI.⁴

At present, the most prospective use cases for AI in the legal profession involve tasks traditionally undertaken by junior lawyers, clerks or paralegals. For example: 

• Legal research: Unfamiliar with an area of law? Wondering how courts have previously decided cases with legal or factual similarities to yours? Drafting submissions? Depending on the complexity of a given topic, it might take a junior lawyer hours (or days) to work through case law, legislation and legal commentary, and then to summarise the relevant information into a clear and comprehensive work product. With AI, there are claims that such research could take minutes.⁵

• First drafts of correspondence: Junior lawyers often draft the first iteration of emails and letters to clients, opposing lawyers, third parties and courts, which are then reviewed and/or signed by their supervising lawyers. Not only is it a learning opportunity for junior lawyers, and (eventually) a time saver for senior lawyers, but it also benefits clients from a costs perspective to have the work front-loaded by junior – and therefore cheaper – fee earners. However, according to a survey by Bloomberg Law in 2023, 58% of the 450 legal professionals interviewed in the US stated that they are now using generative AI to draft, or provide a template, for such communications.⁶

• Due diligence: The review and analysis of contracts, policies, corporate filings and other documents to identify key information and flag potential risks and liabilities is a task typically assigned to junior lawyers. It might take between two to four hours to review 100 pages of material. CoCounsel, on the other hand, can reportedly analyse hundreds of pages and answer complex questions about the material in minutes.⁷

• Discovery: Depending on the size and complexity of a case (and various other factors, such as the number of parties, scope of a dispute and/or requests for further and better discovery), it can take weeks, if not months, to review and categorise documents during the discovery phase, even with the assistance of technology. AI discovery tools, such as aiR for Review, which may reduce both the time and cost of complying with discovery obligations, are increasingly attractive for law firms and clients.⁸

The above list is not exhaustive, and as legal AI tools continue to develop at a rapid pace, the variety of ways in which they might be used within the legal profession – and our confidence in applying them – will grow. It is unavoidable that AI is, or will be, more efficient and effective at some tasks that junior lawyers. Human beings are capable of many things, but processing thousands of documents or search results in seconds is not one of them. 

Adapting Role of a Junior Lawyer

AI will not replace junior lawyers, but it is likely to have a fundamental impact on the nature of the role. After all, if legal AI tools are taking on a growing share of routine, administrative and data-driven tasks that would typically be undertaken by junior lawyers (and clerks and paralegals), where is their time best spent? 

This section discusses two key trade-offs that law firms may face in this transitory period. First, AI’s ability to free up junior lawyers’ time for more complex, engaging work, while also potentially hindering their development of essential legal skills, and second, the potential risks of overreliance on AI without sufficient supervision and verification. 

Division of Labour

On the surface, the automation of repetitive or time-consuming tasks undertaken by junior lawyers, such as case summaries, event chronologies and due diligence, is a win-win: clients save on costs, and junior lawyers are freed up to tackle more complex, engaging work. 

Some argue, however, from a technical and commercial perspective, that these seemingly mundane tasks are necessary steps in building foundational skills, such as learning how to analyse and apply the law, interpret legislation and contracts and understand legal procedure. Slavish reliance on AI and automation may inadvertently hinder junior lawyers’ development of these (traditionally) essential legal skills, and subsequently their ability to effectively oversee AI-driven processes and output.⁹

At the same time, the sheer volume of material that junior lawyers handle today has grown exponentially compared to pre-internet generations. Gone are the days of sifting through hard copy files and pouring over reference materials and reported volumes of cases at the library. With electronic databases containing (in some cases) millions of documents, and the universe of case law, legislative material and commentary at our fingertips, junior lawyers are often under significant time pressure to produce complex, thorough and lengthy pieces of work.¹⁰  If AI tools can halve the time taken to perform such tasks, freeing up junior lawyers to add value in other ways, such as drafting advice and legal submissions, case management, developing client relationships, business development and strategic planning,¹¹  doesn’t it make sense to take advantage of AI’s capabilities? 

One view is that such tasks have traditionally been the purview of senior lawyers, not just because junior lawyers lack the time and capacity to do them, but because seniors have an extra layer of experience and insight, developed over years of practice, which some argue inform these tasks.  

Nevertheless, as law firms incorporate AI into legal practice, the division of labour will inevitably shift. It will be important to upskill junior lawyers in these areas, albeit recognising that their initial contributions to strategic decisions or client relationships may be developmental, rather than immediately transformative. The goal should be to develop firm leaders and industry specialists of the future, rather than tomorrow^.12

Risk of Overreliance

AI might be able to automate or streamline certain legal tasks, but the output produced, whether by GPAI tools or specific legal AI tools, is not reliable or foolproof without close supervision and verification.

One key example is the occurrence of AI hallucinations – i.e. responses generated by AI tools that contain false and misleading information presented as fact. In the case of GPAI tools, such as the well-known large language model (LLM) ChatGPT, some law firms have reported that AI “got the law wrong, misunderstood the question being asked, ‘hallucinated’ citations and confused Australian case law with that of other jurisdictions such as the US and the UK.”¹³ As for legal AI tools, a Stanford University study found that, although they hallucinated less than GPAI tools, they still “make up false or misleading information between 17 per cent and 33 per cent of the time” when used for legal research.¹⁴

Remember the New York personal injury lawyer who was fined for misusing ChatGPT after he submitted a court filing with fake case citations last year?[15] He’s not the only one. In October 2024, a Melbourne lawyer was referred to the Victorian legal complaints body after filing a list of fake case citations in a family court matter, which had been prepared using an AI tool on legal software.¹⁶

Last year, the Chief Justice of the NSW Supreme Court cautioned that AI has the potential to “encourage or feed laziness in research and analysis and a loss of essential skills and critical thinking” and reminded the profession that it “cannot replace the essential function of lawyers in resolving fundamentally human disputes and controversies”.¹⁷ Following his Honour’s comments, the use of AI in the preparation of affidavits, witness statements or other evidentiary material filed with the NSW Supreme Court, and in the drafting and editing of judgments, among other things, was prohibited.¹⁸

A lawyer is responsible for knowing when they can and can’t use AI, and must take ownership of the work that they produce internally, for clients or the court. Where the use of AI is permitted, neglecting to supervise or verify the AI output may lead to costly consequences for law firms and their clients. While junior lawyers are not, in many cases, the final decision-makers, they still need to be aware of the risks and rules around AI and own their step in the process. Blaming false, misleading or inaccurate information on AI is not an option. 

Finding the Right Balance

There are a myriad ways that the use of AI in law firms may be advantageous and/or disadvantageous to the skills development of junior lawyers. Having identified some of these, the question becomes: What should we do about it? How do we make the most of AI’s capabilities and time savings without sacrificing (or neglecting) the development of junior lawyers’ essential skills? 

The balance lies in collaboration. AI should not be seen as a replacement for junior lawyers, nor as an alternative for performing legal tasks without supervision or review. Rather, AI should be used to augment a junior lawyers’ developing capabilities and work output. 

The onus of ensuring that junior lawyers don’t fall behind as a result of AI automation or augmentation is not theirs alone to bear. Law firms and academic institutions share the responsibility for providing junior lawyers (and, before that, law students) with opportunities to develop both essential legal skills that enable them to do the work and future-focused tech skills that will help them do it better. 

At the end of the day, AI might help us do our work quicker, but our human qualities –  empathy, emotional intelligence, critical thinking, and the ability to connect on a personal level – remain invaluable. So long as junior lawyers are given the opportunity to nurture these qualities, AI should not be seen as a threat to their roles but as a tool to enhance their potential. Lawyers who embrace AI’s possibilities will outlast those who don’t. 

The authors acknowledge that a shorter version of this article was originally published by Lawyers Weekly on 3 March 2025.¹⁹