ADOPTING AI: IT’S NOT IF – IT’S HOW

In an article for South Carolina Lawyer, Kim, Lahey & Killough firm founder Doug Kim discusses the risks and rewards of the responsible use of AI for attorneys and law firms. While most in the legal profession bring a cautious approach to the use of AI, this article expounds on the positive benefits of the use of the transformative technology for both attorneys and their clients.

Will AI replace lawyers? No. AI will not replace lawyers but instead will serve as a powerful tool to enhance their capabilities, much like how a high-quality instrument amplifies a professional musician’s training and expertise. AI cannot replace the essential human elements of legal practice. Complex legal matters require nuanced judgment, ethical reasoning, and the personal connection that only human lawyers can provide. AI lacks the ability to make decisions based on empathy, interpret circumstances subjectively, or understand the professional responsibilities that guide legal practice. These human elements remain irreplaceable and essential to effective legal representation.

What AI can do is make lawyers more effective and efficient.

When properly implemented by skilled professionals, AI can automate routine tasks such as legal research, document review, and initial drafting, allowing attorneys to focus on complex strategic work that requires human judgment, ethical reasoning, and personal client relationships.

However, skilled implementation is key as recent court cases involving fabricated AI-generated citations underscore the critical importance of proper training, verification and human oversight.

For clients, responsible AI adoption translates to more efficient service delivery, reduced costs, and higher quality representation, as smaller firms can compete more effectively with larger ones by leveraging technology.

The legal profession is moving toward a future where firms that embrace AI responsibly will provide superior service at competitive rates, while those that resist this evolution risk falling behind.

Ultimately, AI represents not a threat to the legal profession but an opportunity to deliver better outcomes more efficiently, provided it’s implemented with the skill, oversight, and ethical standards that effective legal practice demands.

Read the entire article below, or online here on pages 32-35 and 45 of South Carolina Lawyer.

 

ADOPTING AI: IT’S NOT IF – IT’S HOW

by Doug Kim for South Carolina Lawyer Magazine, a publication of the South Carolina Bar.

Last year, Thomson Reuters released a report[1] stating that 82% of lawyers surveyed said AI can be applied to legal work, but only 51% said that it should. That leads to two questions. What can technically do and what are its risks and rewards? Second, how should lawyers use AI? Part 1 of this article addresses the first question. Part 2, which will appear in an upcoming issue of South Carolina Lawyer, provides my views on the latter question.

Circulating among the profession is the fear that AI will just replace lawyers entirely. This is simply not true. Medium and high complex legal tasks require nuanced understanding, interpretation, and application of the law to specific facts, which are skills that AI lacks. AI cannot make judgment calls based on ethical considerations, empathy, and subjective interpretations of circumstances; these human aspects of legal practice cannot be replicated. Further, AI cannot replace the personal connection that lawyers establish with their clients. Importantly, lawyers are bound by rules of professional responsibility, including confidentiality, competent representation, and client advocacy. AI lacks the ability to honor – or indeed even comprehend – these obligations.

If AI will not replace lawyers, then what good is it? AI is a technology that underlies the next generation of tools to be used in the legal profession. Like all tools, the education, experience, and skill of the user makes a dramatic impact on the effectiveness and quality of the work product created with the tool. For example, a professional musician would benefit greatly from the dramatically improved performance of a Jackson Soloist guitar as the Jackson provides better tone, performance, speed and overall playability. A novice would unlikely know the difference and the performance of the novice would not be materially improved.

According to one McKinsey report, skilled users can leverage AI to replace up to 30% of their work, leading to increased efficiency and productivity – the operative work being “skilled.” AI is not a substitute for education, experience or skill. While AI can accelerate education and skill development in the legal process by automating routine tasks and allowing legal professionals to focus on more complex and strategic aspects of their work, it is not a substitute. Without this requisite skill, the AI tool results in little benefit and can even provide a determent.

The real world provides examples of this risk. Just ask the lawyers that, despite clear judicial warnings and sanctions, continue to submit AI-generated court documents with fabricated (also called “hallucinatory”) content. For example, in Kohls v. Ellison, 2025 U.S. Dist. LEXIS 4928 * (D. Minn. Jan. 25, 2025), the expert witness admitted that his declaration inadvertently included citations to two non-existent academic articles, and incorrectly cited the authors of a third article. These errors apparently originated from use of GPT in drafting his declaration and the expert failed to verify these fake citations to academic articles before including them in his declaration.

Indeed, the problem has affected how judges take testimony. In In re Matter of Weber, 2024 NY Slip Op 24258 (N.Y. Surrogates Ct. Oct. 10, 2024), the trial court refused to accept as accurate without more back-up calculations performed by artificial intelligence. The court stated that due to the nature of the rapid evolution of artificial intelligence and its inherent reliability issues, before evidence which has been generated by an artificial intelligence product or system is introduced, counsel has an affirmative duty to disclose the use of artificial intelligence, and the evidence sought to be admitted should properly be subject to an admissibility hearing prior to its admission.

Properly used, AI tools can help lawyers automate manual processes and work more efficiently. For example, when using the AI tool for patent application drafting, more time should be spent at the beginning on the drawings of a patent application and generating a bullet point documents of key features of the invention’s structure and function that are novel and non-obvious (patentability requirements). This shift changes how patent lawyers invest their time, but they do it to more effectively use the AI tool.

Similarly, AI-powered case law research tools can help lawyers more easily find leading case law and guiding legal principles than old-fashioned Boolean searches, but only if lawyers can move away from that conceptual framework. Otherwise, a search term mentality fails to take best advantage of the tool.

Just as important – and unlike traditional computer-aided research – AI can suggest the best language to support arguments. Indeed, AI can also help lawyers quickly produce initial drafts of motions, legal briefs, contracts, and settlement agreements, which in the past have been time intensive. Improper use of AI, however, will cause the output to be either less useful or even counterproductive.

More generally, AI will change how legal services are provided. Routine tasks, such as legal research, document review, and billing can become automated, allowing lawyers to focus on more complex and strategic aspects of their work, especially those that require critical thinking based upon the facts and circumstances of the parties’ case and desired outcome. AI can assist in predicting case outcomes, advising clients, and developing informed strategies. All that requires, however, that lawyers re-order what they do and how they do it. If a lawyer traditionally spends 30% of her time in complex cases reviewing documents, it will require substantial willpower to pivot to a profoundly different allocation of time.

This will also impact the profession as a whole. While not replacing lawyers, AI will create a demand for lawyers with specialized skills in areas that require complex problem-solving, critical thinking, negotiation, and dispute resolution. Substantively, with AI fulfilling some of the more mundane tasks and day-to-day needs, the need for “specialized” areas (e.g., tax, securities, environmental, employment, intellectual property, complex litigation, and the like) will increase. But that also means that lawyers who have made a good living out of routine tasks will face an increasing challenge.

I experienced a movement years ago when I was a computer programmer prior to becoming a lawyer. The same positions were being put forward that AI was going to replace computer engineers. This did not happen. In computer science, AI, especially large language models like Chat-GPT, proved to be just another tool in the toolbox and made experienced programmers more efficient and accurate, enhanced code quality, reduced development costs, reduced time to market, improved quality control and generally made for a better work-product; especially experienced programmers. Properly implemented, AI will do the same for the legal profession. AI will not replace the lawyer.[2] However, attorneys that adopt AI will replace those that do not.

Further, the smaller law firm will become more competitive through leveraging technology just as we saw with online research, communications tools, timekeeping and billing, and practice and case management software. By automating certain tasks, smaller firms can take on more clients without needing additional resources and can maximize profitability while delivering improved legal services to clients. This levels the playing field and allows smaller firms to compete with the larger ones so that the dreaded overhead of the larger firm will become even more of a burden that results in higher fees and less value to the client.

This is not to say that the transition will not be bumpy. While adopting technology eventually reduces overhead and provides more value to clients, the legal profession has always faced challenges when it comes to innovation and technology. Further, most of the bar was licensed prior to the release of the iPhone.[3] This “slow to adopt” challenge stems from the reality of technical competence, or lack thereof, and the resistance to change which stems from being severely risk averse.

In 2017, the Association of Corporate Counsel published A Seat at the Head of the Table: From Lawyer To CEO? and stated, “Risk aversion. Most law firm cultures are built around a strong bias to risk aversion. However, … lawyers must expand their perception from ‘no’ to ‘yes,’ even if the next question is “how?”[4] While this publication was discussing lawyers transitioning to CEOs, it is applicable to the current discussion of whether to “use” AI. It is not a matter of yes or no, but rather when and how. While lawyers didn’t go to law school to be computer programmers, the technical requirements of modern legal practice cannot be ignored.[5] By analogy, the craftsman who used hand tools for many decades may be slow to adopt power tools for the same reasons that lawyers will be slow to adopt AI: habit, inertia, fear of change, inflexibility, etc. Yet history shows that those who adopted power tools surpassed those who did not in short order.

Disruptive technologies are just that – disruptive – and seeing the coming changes and being open to this disruption in the legal industry is critical. Disruption often comes without warning and is not anticipated by the masses. A quote often attributed to Henry Ford is, “If I had asked people what they wanted, they would have said faster horses.” Adapting to disruptive technologies is important to success and potentially the survival of the modern law firm.

AI is growing faster in its development than prior disruptive technologies. This creates both risks which are manageable and opportunities which will further the profession. These risks can be effectively managed with a few simple guiding principles:

• Understand how the AI works, including its training model and limitations.
• Implement AI to a specific application and avoid general, open AI tools.
• Make sure you are always in control of the technology and be accountable for its use (avoid the chatbot).
• Try to only use specific AI that has private, dedicated connections to large language models (as opposed to public data lakes like used by ChatGPT).
• Use AI tools that do not use your input to train with the AI engine and keep input confidential.
• Always have a live person review the content and work product.

The power of AI and the benefits that it can bring will create a separation in the legal profession, split into those that properly adopt AI and those that do not. The split will manifest in lawyers’ ability to deliver higher quality legal services and work product, the faster delivery of work product, and less costs; each of which impact referrals, reputation, and client satisfaction. Reuters reports that clients expect their lawyer to lead the charge in leveraging AI for the client’s benefit.[6] Clients will expect law firms to use AI due to the potential reduction in cost and time savings. The law firm that responds to these demands will have a significant advantage over those that do not. Adopting AI will also lead to changes in billing practices, and we will see more firms move to alternative fee arrangements such as a flat fee billing arrangement.[7]

Doug Kim, a Physics major from Davidson College, began his professional career as a computer programmer and software engineer. His intellectual property career began in 1998 when he combined his business experience with his legal education and was involved with enforcing a client’s patent against multiple infringers. Since then, Doug has created a well-rounded IP practice that provides legal solutions and strategies tailored to each client from multinational corporations to start-ups.

[1] New report on ChatGPT & generative AI in law firms shows opportunities abound even as concerns persist, Thomson Reuters, April 17, 2023.

[2] In all fairness, we should see a reduction in the needs for support staff and junior associates. These roles typically involve administrative, repetitive tasks that AI can now perform with greater efficiency and cost effectiveness.

[3] The iPhone was first released June 29, 2007.

[4] A Seat at the Head of the Table: From Lawyer To CEO, by David Felicissimo, Association of Corporate Counsel, 2017.

[5] See CLE requirements in North Carolina for Lawyers, NC Continuing Legal Education (showing that lawyers in North Carolina must complete one hour of CLE in Technology every two years.) https://www.nccle.org/for-lawyers/requirements/renewing-lawyers/

[6] AI: the new legal powerhouse — why lawyers should befriend the machine to stay ahead by Jeremy Glaser and Sharzaad Borna, Thomson Reuters

[7] I appreciate the natural financial friction where the lawyer will pay for AI just to be able to charge less. However, while beyond the scope of this article, AI will have an impact on the adoption of alternative fee arrangements. I have been providing flat fee arrangements for decades and trust me, there is nothing to fear.