Legal work can require hours upon hours of research. Attorneys and their aides scour through previous legal cases, statutes and applicable regulations to find the information needed to help their client win their case. Over the years, various tools have emerged to help increase the efficiency of this process but few offer as much potential as artificial intelligence.
How is AI changing the practice of law?
A recent report by Thomson Reuters notes that more than 50% of attorneys were seeing a return on investment (ROI) from use of AI within their organizations in 2025. More attorneys are embracing the use of AI to help with document review, research and drafting contracts, potentially saving lawyers hundreds of hours of work every year. This does more than save lawyers time, it can also translate to lower bills for clients.
Will use of AI tools be expected within legal work?
With all of these benefits, it is no surprise that a New York federal judge recently raised a pointed question: could a lawyer’s refusal to use AI become a malpractice issue? The question is relevant to this moment in time, as courts and clients push toward measurable efficiency. Attorneys must balance this expectation with the fact that professional responsibility rules still center on competence, supervision and confidentiality. AI use sits at the intersection, creating new arguments on both sides of what it means to be a reasonable practitioner of the law.
It is important to note that the use of AI in the legal world is not all good news – many of us likely remember the attorney on a video call during the pandemic that could not shut off the AI which turned his image into that of a cat. Although relatively humorous, other issues are much more dangerous. From hallucinations to bias, AI can also pose serious problems and setbacks to the pursuit of justice. A reliance on misstated facts from precedent that does not exist can cause serious damage to a client’s case.
How do courts rule on these matters?
At this time, courts and the legal practice continue to require human oversight. The lawyer behind the legal representation is responsible for the service provided to clients — not an AI tool. Judges have imposed sanctions for false citations tied to generative AI. Those orders reinforce a simple rule: attorneys remain responsible for the work product, regardless of tool choice.
It appears the debate is likely to shift. No longer will the question be whether AI belongs in legal practice. Instead, it will be about when nonuse becomes unreasonable and reckless as well as how lawyers prove they exercised independent judgment.