(Header Image Source: Euractive)
The legal profession has been a cornerstone of human civilization for over two millennia. From the Ancient Roman Republic to today, the storied history of the lawyer has been one of evolution and constant change. But perhaps nothing may ever pose as significant an impact on the legal profession as the current advent of artificial intelligence.
What is Artificial Intelligence?
Artificial intelligence (AI) is a fast-growing phenomenon with ever-expanding applications in our everyday lives. While many have excitedly jumped on the AI bandwagon, few have stopped to ask, “What truly is artificial intelligence?”
So I asked AI itself.
ChatGPT gave a very helpful response:
“AI is a broad field of computer science that is focused on creating smart machines capable of performing tasks that typically require human intelligence. It encompasses a variety of technologies and methods, including machine learning where computers are given the ability to learn and improve from experience without being explicitly programmed.”
Rather aptly, this definition aligns with that of John McCarthy, the late pioneer of the AI discipline, who described AI as “the science and engineering of making intelligent machines.”
As AI has continued to progress over the years, it has found itself increasingly useful in numerous fields, such as medicine, farming and even art. The legal profession is no exception. According to a 2023 study by researchers at Princeton, the University of Pennsylvania and New York University, the legal field was the industry most exposed to new AI language modelling, which involves the use of algorithms to interpret, predict and generate human language — much like ChatGPT.
So, what does this mean for the legal profession?
The Benefits of Artificial Intelligence
First, we begin with the obvious benefit that stems from AI and technology in general: automation. Economists at Goldman Sachs found that 44% of legal work could be automated. Through automation, AI will usher in a new era of efficiency in legal practice by streamlining the tedious, time-intensive tasks that have traditionally been pawned off to junior associates and paralegals. The dreaded process of combing through dense legal documents to pinpoint key information will be accelerated by advanced software, allowing legal research to be performed with greater speed and accuracy. This technological shift will not only reduce legal costs by expediting time-draining processes but also redirect valuable resources and labour to more sophisticated roles within legal practice.
Furthermore, the increased productivity will enable firms and practising lawyers to handle more cases more rapidly and take on many new cases. This, in combination with reduced legal prices, would have the added benefit of making legal representation more accessible for everyone, especially those of lower socioeconomic status.
But it would be naive to assume that increased efficiency is the extent to which AI may automate the legal industry. Indeed, it is alluring to think of ways in which complex, ever-improving AI language models might someday breed some ultimate legal algorithm, a means to overcome the very human limitation of our modern legal system: the lack of strict objectivity, particularly in how court cases are decided.
The subjective application of legal standards of proof by various judges and juries often results in inconsistent decisions and wrongful convictions. With AI, there is a potential future where deciding legal cases is transformed into an exact science.
In civil cases, a future algorithm could analyze case specifics and render a decision using a precise 50% threshold, aligning with the required civil standard of proof of a “balance of probabilities” (i.e. more likely than not). Likewise, in criminal cases where the standard is “beyond a reasonable doubt,” the algorithm would judge based on a higher threshold of proof, potentially around 95%.
This AI-driven approach would revolutionize the predictability and accuracy of legal outcomes, significantly bolstering public confidence in the justice system. Although this transformative change might be a long way off, it would provide the most significant impact the legal system will likely ever see.
The Dark Side of Artificial Intelligence
But it isn’t all sunshine and rainbows — with all the benefits that come with AI also come many disadvantages. For example, considering 44% of legal work can be automated (as mentioned before), many lower-tiered jobs could be replaced by AI, especially those of paralegals and legal assistants.
A 2016 study from researchers at the University of North Carolina School of Law and Massachusetts Institute of Technology revealed that AI’s primary contribution to increased productivity lies in general document management. The study noted that most of these tasks were traditionally handled by paralegals or clerical staff, indicating that these groups would be most affected by AI labour displacement.
However, AI would still have a sizable impact on lawyer employment. The same study found that the greatest impact would be in the area of document review, which involves the process of examining documents for litigation or government investigations.
Furthermore, the ethical issues surrounding AI cannot be overlooked. Given that AI relies on human algorithms and data, it can incorporate implicit biases or reflect historical inequities, leading to concerns over discrimination.
For example, ProPublica’s investigative study on a criminal justice algorithm used in Florida revealed significant disparities. The report highlighted that Black individuals were nearly twice as likely as White individuals to be incorrectly labelled at a high risk of recidivism. Conversely, White individuals were more likely than Black individuals to be labelled “lower risk” but go on to commit further crimes. These issues must be addressed before AI can be safely integrated into professional contexts.
Many also raise concerns about the idea of AI replacing humans in jury trials. Stratton Horres, retired senior lawyer at Wilson Elser, maintains that the intricate nature of legal proceedings often demands moral deliberation and an understanding of the nuances of human behaviour. Horres points out that jurors possess distinct attributes like empathy, common sense and the capacity to assess witness credibility through non-verbal signals — qualities that AI systems cannot replicate. As such, the role of AI in jury trials should be limited to supporting and enhancing the decision-making of humans, not replacing it.
What will the legal field look like in years to come?
During our interview with Fadi Matthew Kazandji, a leading criminal and family lawyer in Toronto, he pointed out that AI could quickly benefit the public, especially self-represented litigants, as they can more easily find relevant answers to their legal questions without needing to spend money on a lawyer.
However, Kazandji believes we must be careful of how market forces will affect access to AI in the future. Currently, the basic services of ChatGPT and many of its competitors are available for free, but include premium subscriptions with enhanced features. According to Kazandji, “It is possible that as AI models evolve, their owners and creators will crowd out competition and impose near-insurmountable barriers to entry. Then, they can leverage their market power to charge users far more. This can create a two-tiered legal system where only the wealthiest clients can afford legal services that incorporate the most advanced AI.”
Moreover, Ben Allgrove, chief innovation officer at Baker McKenzie, believes that AI will “force everyone in the profession, from paralegals to $1,000-an-hour partners, to move up the skills ladder to stay ahead of the technology.” The role of humans will be to increasingly focus on cultivating industry experience and establishing trusted relationships with clients — areas where AI falls short.
Ultimately, AI has the distant potential to quantify the unquantifiable and bring objectivity to court cases, transforming the legal field far beyond current expectations. But in the foreseeable future, it appears that AI will primarily reduce the functions of paralegals and legal assistants, with relatively moderate impacts on lawyers.
Yet, this change is still some years away. For most of those currently employed in the legal sector, there is little cause for concern — at least for now. Case in point, consider the story of an American lawyer who used ChatGPT in his court case. The chatbot “hallucinated” fake cases, which the lawyer then cited in his briefs. As a result, he was fined $5,000 by the judge and lost the case.
The moral of the story? Ironically, the only significant threat to legal jobs from AI so far appears to be over-dependence on it. This serves as a cautionary tale, suggesting that we must embrace AI not as a reliable consultant to which we may outsource our duties, but instead as an eager yet disorganized new assistant in need of supervision.