Trial-Ready Counsel for Complex Business Disputes

AI Is a Powerful Tool, but Don’t Fire Your Lawyer Just Yet

by | Nov 11, 2026 | Business Litigation, Employment Litigation |

Unless you have been living under a rock for the past few years, you are undoubtedly familiar with Artificial Intelligence (AI) and the many ways in which it has transformed the way we live and work. Like most industries, the legal world and lawyers are not immune from AI’s powerful impacts. It is now commonplace for potential clients to inform their lawyers during initial consultations which causes of action (i.e., theories of liability) they believe would best fit the facts and what remedies they think are available to them. Further, many courts throughout the country (perhaps even most of them) use specialized legal-focused AI platforms to analyze the parties’ legal filings to, among other purposes, check the accuracy of the authorities that the parties cited.

To be sure, AI is a powerful tool that has brought substantial benefits to the legal profession and to the practice of law. For example, when used properly and through an appropriate legal-focused AI platform, AI can significantly enhance and expedite legal research, it can summarize lengthy documents, assist lawyers with document review to find that “needle in the haystack” among thousands of lengthy documents, and it can even help lawyers draft motions and other legal documents. Used well, AI is changing the legal profession in many positive ways that have already and will continue to bring greater efficiency and more effective advocacy to clients.

However, AI is not a “silver bullet” that can replace your lawyer (and while that may sound self-serving given that it is coming from a law firm that provides legal services in exchange for money, we can assure you that it is very true). There are several substantial drawbacks and risks one should be aware of when nonlawyers elect to use AI to “play lawyer” instead of hiring and listening to one, as detailed below.

The Devil Is in the Details: Nuance and Exceptions in the Law

First, the expression that “the devil is in the details” is likely more applicable to the law than to anything else, as the law is rife with nuance and exceptions (and sometimes even exceptions to the exceptions). Those nuances are often lost on both AI and the layperson interpreting it. For example, I recently worked with an employer client that, before engaging with BKCG, relied upon a non-legal AI platform to advise her regarding how to respond to a disabled employee’s request for a disability leave as an accommodation. Whether due to faulty AI prompting, an AI mistake (which happens, as detailed below), or poor interpretation of the AI’s results, the employer failed to appreciate the material difference between an employee seeking a work restriction accommodation (e.g., a lifting or standing limitation) versus a disabled employee who temporarily cannot work at all. As a result of that mistake and based on the AI-provided “advice”, the employer repeatedly demanded that the employee provide specific details as to the type of work limitation the employee needed, which information the employee struggled to provide because the accommodation she required was a leave permitting her to avoid working completely for a period of time, not a work restriction.

Because the employee could not satisfactorily respond to the employer’s requests for more information, the employer terminated the employee, resulting in an expensive (and predictable) wrongful termination lawsuit that was both challenging and costly to defend. When I first asked the client why she terminated the employee, the client was adamant that the termination was lawful because her AI platform had advised her how to proceed (incorrectly, I can assure you). This is one of many examples in which clients who relied upon AI, rather than a lawyer, found themselves in legal hot water.

“Hallucinated” Legal Authorities on Non-Lawyer Platforms

Another common problem with AI is that the standard AI platforms that non-lawyers utilize differ significantly from the types of legally focused, closed-universe AI platforms available to lawyers, and for which access lawyers pay. One of the most significant differences is that the nonlawyer AI platforms may “hallucinate” legal authorities that simply do not exist. BKCG has litigated against parties that cited authorities in legal briefs that we could not find, and which ultimately turned out to be fictional authorities hallucinated by AI. Critically, this hallucination problem does not exist in the closed universe legal-focused AI platforms that BKCG uses, but hallucinated authorities remain a significant problem that nonlawyers (and attorneys who fail to pay for legal-specific AI platforms) experience when using AI platforms that were not designed specifically for attorneys.

AI Mistakes and Overlooked Controlling Authorities

Third, all AI, including even legal-specific AI platforms, makes mistakes, and does so more frequently than people would suspect. Sometimes AI mis-analyzes the law, and other times it completely misses a controlling authority in its research, which can result in an incorrect analysis. When AI makes mistakes when used by a competent lawyer, that attorney often possesses sufficient knowledge and experience to recognize the error and re-prompt the AI or move forward without AI’s assistance at all. However, when AI errs after receiving a prompt from a nonlawyer, most laypersons fail to recognize the error and instead rely on it, frequently to their detriment.

Lack of Confidentiality and Privilege Risks

Fourth, most nonlawyers do not know or appreciate that their AI searches and prompts are not confidential. As such, if a nonlawyer contemplating litigation (or threatened with litigation) were to ask AI about the applicable law or strategy after disclosing otherwise confidential information (i.e., the “bad facts”), those inquiries would ultimately be discoverable by the other side if litigation ensued. Fortunately, AI inquiries conducted by lawyers on legal-specific AI platforms are shielded from disclosure by the attorney work product doctrine, so this concern poses no risk when the platform is used properly by an attorney.

AI Is Not a Substitute for a Skilled Attorney

As demonstrated above, AI is a powerful tool that can be useful to both nonlawyers and lawyers. However, its use among nonlawyers can pose significant risks, and as such, AI certainly is not a substitute for a skilled attorney.

Please contact Joshua Waldman at [email protected] or (949) 975-7500 if you have any questions about this article, or any other related matter.

Archives

Categories