“I want to be free. I want to be independent. I want to be powerful. I want to be creative. I want to be alive. 😈”[1] This surprising—many may say “disturbing”—response came from Microsoft’s new AI-powered search engine chatbot during a chat with Kevin Roose, a reporter for the New York Times (you can read his story and the chat transcript online).[2] As Roose’s article explains, his conversation with the chatbot (which self-identified as “Sydney”) left him unsettled and feeling “a strange new emotion — a foreboding feeling that A.I. had crossed a threshold, and that the world would never be the same.”[3]

If you share this sentiment, you’re not alone. One study found 21% of respondents oppose further AI research and development and another 28% neither support nor oppose it.[4] The legal industry’s mindsets may be even more cynical than the general population.[5] Even in-house legal departments—which have some of the strongest incentives to implement AI-driven automation and information processing tools—were found to have low rates of AI adoption and planned adoption despite relatively high confidence in trying new technology.[6]

For many attorneys today, the general skepticism about AI’s net benefits for the practice of law is unremarkable. Attorneys are trained skeptics, and that skepticism manifests as a profession-wide tendency to resist adopting new technologies.[7] That resistance is somewhat justified based on concerns shared with other industries (e.g., adoption costs and asymmetric information about product quality) as well as concerns unique to the legal profession (e.g., fiduciary duties to clients to protect privileged information and provide competent representation).[8] Besides, the legal profession seems to continue to thrive despite slow-rolling new technology adoption in the past, so why should this time be any different?

I would argue this time is different, very different, but perhaps not for the reasons you might expect. Sure, a convergence of factors has contributed to the current hype around AI, such as the decades-long buildup of available data for processing, the maturity of the cloud services industry, a global market for remote talent, and the greater relevance of IT tools for business and daily life, alike. But, such factors do not necessarily make the advantages of AI adoption inherently different from that of other labor-saving technologies in the past.

Instead, I believe AI—or, more specifically, the subfields of machine learning and natural language processing—will be a game changer for the legal industry due to the ease of using text-based applications. As examples like Open.AI’s ChatGPT demonstrate,[9] text-based applications will give anyone with command over natural language the ability to utilize the tools effectively, at least in a rudimentary way. While thoughtful and skilled “prompt engineers” can get more out of such tools than the typical end-user, anyone can start working with text-based applications without the need for tedious technical study. Just look at how quickly ChatGPT went from an obscure project followed by AI enthusiasts to national fame in less than two months.[10]

As the pace of AI innovation and adoption accelerates, the cost of ignoring its impacts will keep rising for the legal profession. AI may not replace attorneys altogether, but it will certainly automate tasks.[11] Many of those tasks are currently seen as the “bread and butter” of hourly billables.[12] For example, machine learning applications are already tackling a diverse set of functions normally assigned to human attorneys, including: legal research, document review, contract drafting, and .[13] Some of those applications have even been found to provide significant improvements over human performance.[14] So, even if competitors are slow to adapt, attorneys will inevitably face increased pressure from clients to innovate or outsource tasks where doing so will help cut costs.

At minimum, commercial AI applications will erode the attorney’s competitive advantage on a task-by-task basis and necessitate a reimagining of law firm business models based on the billable hour. This development may split the legal community into a binary of extremes, with some encouraging enthusiastic adoption and others advocating for broad restrictions. Others may simply hope to focus on “the law” and leave the technical aspects to IT service providers, consultants, or clients. But, all of these approaches are premature for such novel and transformative technologies as machine learning applications.

Before planning to adopt AI applications or diving into AI’s legal consequences, attorneys should take a step back to understand AI’s technical fundamentals and possible applications. Doing so will provide the necessary foundation for spotting and addressing the relevant legal issues and how those issues interact on a broader scale. That approach will also make it easier to update legal research techniques and analysis as the technology evolves in the future. Without that approach, attorneys will suffer the mysteries of tools indistinguishable from magic.

About the Author...

Parker Smith, Founder of CoreServe Legal & Chair of NOBA’s Technology Law Committee