The potential and drawbacks of using artificial intelligence in the legal field

Extraordinary advantages and harrowing disadvantages of employing AI in the practice of law

2023 November

When The Matrix movie came out in 1999, I sat wide-eyed in the movie theatre, thinking this radical, surrealistic, computer- simulated world will never be something I’ll see in my lifetime. Even with all the hype Y2K was carrying at the time, the notion of artificial intelligence, better known as “AI” seemed way too cool, and utterly science fiction. How can a robot understand human talk and carry out the requested task with just a few vague commands? Impossible. Big thoughts for a six-year-old, and as it turns out, so very wrong.

Fast-forward over two decades, and there I was, immersed in a labyrinth of legal theories, spending hours researching nuanced concepts, poring over case law and court opinions thicker than a law school textbook, on what seemed like a never-ending quest to find the elusive answer to my particular question. When I finally surfaced, I realized I spent far too long for my clients’ (and my own) liking just to conclude my research inquiry with, “Well, it depends.”

Enter the game-changer: AI, which has been transforming many sectors, the legal field being no exception. Legal research felt like pulling teeth. But, thanks to the development of legal-based AI programs, I’ve finally stumbled upon the magic – and horrors – that artificial intelligence can offer myself and my clients, including rapid-fire direct responses to legal inquiries in just a few minutes, and even fewer clicks.

Research projects that once aged me in hours are now zipped up in minutes, citations and statutes intact. Drafting and reviewing documents? Lightning fast. Now, I’m a legal wizard, doing it all – informed, efficient, competitive. Even peaceful, sometimes.

This is not to say that AI does not have its downsides. We have all heard the horror stories: AI reciting incorrect case law, botching legal interpretations, inventing cases, or quoting outdated authorities. We can’t risk getting caught with our pants down, relying solely on AI. AI systems are programmed based on the data they are trained on, and if there are errors, biases, or inaccuracies in that data, it can lead to blunders, sanctions, dismissals, and utter chaos for its users.

Lawyers and legal professionals using AI tools must tread carefully, and exercise caution and diligence when relying on AI-generated research and citations. Double-check what AI spews, cross-reference it with real legal research, and toss in your own smarts. AI is a wingman, not the captain. As such, AI should only be used as a supplementary tool to enhance legal research, rather than relying solely on it.

For California trial lawyers, employing AI tools brings forth a variety of pros and cons which could significantly impact their professional practice and client relationships. It’s like a double-edged sword. It can shape your practice and client bonds, but you’ve got to wield it smartly. It is an ever-evolving landscape that merits close watching and careful utilization in order to be effective for its users. The AI scene is developing further each day; keep an eye on it, make it your ally.

Pros of legal AI software

Efficient legal research

AI can expedite and automate numerous routine tasks that traditionally consume much of a legal professional’s time. Most routine and standard tasks, such as contract analysis, legal research, screening for conflicts, outlining questions for depositions, and voluminous document review can take as little as five minutes to perform, with minimal work (if any) on the user’s end.

Mundane tasks are effectively eliminated from the daily rotation. AI can streamline and quickly perform these tasks, thereby saving attorneys significant research, drafting or review time, allowing them to devote more time to complex, higher-order strategizing, and case-analysis work that requires their expertise, like client engagement and proactive counseling. In effect, much of the administrative work will be done for the user.

There is no elaborate language involved either – AI is intelligent enough to chat intuitively with its user – you speak to AI the same way you would speak to an assistant, without having to worry about hitting every buzz word to yield a proper search.

Research proficiency

AI’s search algorithms are capable of quickly sifting through massive amounts of legal databases, and presenting the researcher with relevant caselaw, statutes, regulations, legal articles and journals about their chosen topic or inquiry. AI’s advanced systems can highlight patterns or correlations that would otherwise be almost impossible for a human to identify on their own, due to the sheer volume of data presented. The user will be presented with available data that they would otherwise not think to look for, or simply don’t have enough time to look into.

Predictive analytics

AI can help lawyers assess and analyze the potential outcomes of their cases based on historical data, aiding in the strategy development of any given case. Predictive analytics can even teach the researcher a thing or two about their case’s potential outcomes, and how to navigate the case, based on the historical data that the researcher may not otherwise be privy to.

 Some people rely on colleagues or mentors to guide them through their legal career, but the presence or guidance of others would not necessarily be required when one can rely on the data and predictive analytics provided by AI. This drives more strategic planning and strengthens the attorney’s legal argument. It can also assist in setting realistic expectations for case outcomes and value with clients.

Cost reduction

By automating daily tasks, legal professionals can potentially reduce the costs of firm operations, legal research, outsourcing tasks and/or hiring persons to perform the tasks. These savings can be passed on to clients, making legal services more accessible to a wider demographic, and more funds for the clients rather than toward their incurred costs. It also frees up the attorney’s workload to take on additional cases that they may otherwise need to turn away due to time constraints.

Consistently updated

By constantly monitoring and analyzing legal databases, statutes, regulations, and court decisions, many AI platforms provide continuous updates about changes in law, ensuring the lawyers have the most up-to-date information. This could help lawyers stay at the forefront of the evolving legal landscape, giving them a competitive advantage. AI can also facilitate ongoing professional development by suggesting relevant courses, articles and resources based on the lawyer’s practice areas, research inquiries, and interests.

Document generation and review

AI can generate legal documents such as contracts, agreements and briefs, faster and with fewer errors, saving its users time, money and stress. It’s like having a digital assistant that gets each task done in under five minutes and is available to help 24/7 (no more overtime!). AI is also capable of document review, which can examine vast amounts of legal documents, contracts, and filings to identify inconsistencies, errors or clauses that may need updating due to changes in the legal landscape.

However, it is critical to note that while the usage of AI in the field of law can seem like the magical, unicorn assistant you always wanted, it also carries potential drawbacks that its users must be privy to before diving in head-first.

Cons of legal AI software

Ethical and data privacy concerns

Using AI software often necessitates uploading sensitive client documents and/or information onto the cloud or web. Storing sensitive legal data in AI systems raises concerns about data security and privacy. This can pose potential risks to client confidentiality if the platform’s security measures are penetrated.

The use of AI raises new ethical quandaries that we are still grappling to understand. Who is responsible when AI makes a critical mistake? Can client- privileged information be secured when shared with AI software? How do the users know if the information they share with AI is truly protected?

Every attorney subject to the American Bar Association rules of professional conduct has sworn an oath to abide by the Rules of Competence in the code of ethics. Comment 8 of Rule 1.1 states, “To maintain the requisite knowledge and skill, a lawyer should keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology, engage in continuing study and education and comply with all continuing legal education requirements to which the lawyer is subject.” Mishandling AI, whether deliberate or not, can be interpreted as a violation of the ethical code, which may cost the user their reputation, employment, and/or license.

Dependence on technology

Overreliance on AI might lead legal professionals to lose their core legal research and writing skills. This may affect their competency in situations where AI tools are unavailable or unsuitable, and often weaken an attorney’s ability and desire to figure some things out for themselves, without reliance on a robot. After all, AI is software, and as with all tech, glitches can occur. Interruptions due to software bugs, network issues, or cyberattacks can disrupt workflow and potentially compromise client data. It’s best to make sure our skills remain sharp, and we never put too much dependence or reliance on any AI software or technology in general.

Limited understanding and depersonalization

The key difference between human beings and AI technology is that AI lacks human intuition and empathy. Law is a profession steeped in person-to-person relationships. While AI can assist in many areas, it cannot substitute human empathy, judgment, and intuition. This may limit its usefulness in comprehending nuanced interpretations of law or understanding the emotional components often involved in trial cases.

In other words, AI does not have a heart, brain, or personality. It merely contains and provides data. Because databases are used to train AI, the AI program becomes vulnerable to the implicit biases of the databases used to train it. In a society where customer service is king, the personal touch that can be offered only by people and natural human emotion is lost, thereby risking a diminishment in attorney-client relationships. Lawyers typically get “repeat” clients for a reason. Part of that is how connected and understood the client feels by their lawyer. Overreliance on AI may get you accurate data, but it can never replace a personal touch, no matter how well AI tries to emulate it.

Lack of contextual understanding

You may find that AI struggles to understand context, leading to responses that are irrelevant or confusing, particularly when discussing complex or nuanced legal topics. The lack of contextual understanding may also lead to incorrect recommendations or authority to support your position as AI tends to understand technical language better than commonly used colloquial language.

Incorrect research or data

One of the biggest horrors of relying on AI in the legal field is when/if AI generates incorrect research or data in response to an inquiry. This disastrous scenario is perfectly demonstrated in the matter of Mata v. Avianca, Inc., 22-CV-1461 (S.D.N.Y.) wherein a lawyer used ChatGPT to assist in briefing a matter he intended to argue before the court. ChatGPT was shown to have created fake cases, raising concerns that AI can and will provide incorrect information.

The Mata attorney asked ChatGPT to provide arguments and find caselaw in support of his position, and then included the AI’s “research” in the court documents. When doubts arose about the legitimacy of these cases, the attorney asked ChatGPT if the cases it provided were real, and the AI confirmed their existence, despite being completely made up.

This catastrophe revealed a tragic misunderstanding of how AI operates. The attorney thought AI was like a powerful search engine. Once it was revealed that the cases it provided did not in fact exist, the lawyers asked the court for leniency, citing their inexperience with AI. Those attorneys went on to be sanctioned $5,000 by the court. I’m sure that conversation with the client was not fun. One can only imagine that the courts will become less and less tolerant of the misuses of AI, and further penalize the offering lawyers – beyond mere monetary sanctions.

Cost and accessibility

Implementing legal AI systems can be expensive, especially in the initial phases, requiring both software and hardware investments, not to mention a costly, recurring membership fee and training. Although competition for legal AI programs is growing, and prices are falling, quality AI software systems can still be quite expensive, which may limit accessibility for smaller firms or solo practitioners.

Job displacement

Task automation may reduce the need for certain legal tasks or positions as a whole, potentially leading to job displacement for some legal professionals. This could lead to diminishment of paralegals, paraprofessionals, research assistants, associate attorneys, and more. There are articles left and right about ChatGPT passing the bar exam in multiple states, including Arizona, Colorado, and New York, with nearly perfect scores, higher than most human bar examinees.

While I don’t think lawyers’ jobs are in complete jeopardy, an unfortunate many can expect to find their jobs replaced by a robot in the near future or find that they can’t offer work product at the speed and efficiency that an AI system can.

People will assume you’re using AI (Even when you’re not)

When I told my colleagues I was writing an article about AI for Advocate, nearly everyone asked me if I was going to ask AI to write the article for me. Despite insisting that I would not, many people assumed I would use AI anyway (come on – is AI this funny?).

Another time, I needed to send my client a strongly worded letter. Before sending it out, I ran the letter by my boss, and he said, “Great job. ChatGPT?” to which I responded, “no, just Gina” which was followed by a righteous dad joke from my boss along the lines of “I thought the ‘G’ in ChatGPT stood for Gina.” I suppose it’s a compliment for good work product to be seen as performed by an emotionless robot, but it has certainly become the norm for legal professionals to expect others to use AI to perform most of their work-related duties.

As we’ve seen, AI holds great potential to revolutionize the legal profession – and to turn it upside down – by offering significant advantages like reducing time-consuming work, improving accuracy, and providing analytical insights. However, it’s not just about dodging bullets. AI can dodge hours of grunt work, enhance precision, and provide insights that even Morpheus would appreciate. But, beware, as these transformations also present challenges related to technological dependency, loss of human touch, ethical conundrums, and budgetary battles.

It’s pivotal that the legal community navigate this newfound path consciously, acknowledging both the potential benefits and drawbacks that could arise from the incorporation of AI in the practice of law. The legal community stands at a crossroads – the red pill (AI) or the blue pill (old-fashioned grunt work).

Before diving headfirst into the AI rabbit hole, it’s essential for California trial lawyers to carefully weigh the scales of justice. Does AI complement your legal prowess, or does it hijack your practice? Make your choice wisely and ensure that it complements your practice rather than hindering it.

So, when contemplating the embrace of legal-based AI, remember, you’re not in The Matrix – you can’t just swallow a pill and hope for the best. Draft, research, and proof your work as if the State Bar were watching. Ignorance in this scenario is anything but bliss.