
Please welcome back Lila Warren, a frequent contributor to the 21st Century Tech Blog. In this article, she describes how artificial intelligence (AI) is changing the nature of legal practice. Will we no longer need lawyers?
Recently, I came across a database of AI hallucination cases. As of November 17, 2026, it indicates 2,312 cases so far, of which 1,894 were pure AI fabrications, 977 misrepresentations, and 613 false quotes.
More than two-thirds of the cases were from the U.S. Three of them were from U.S. government lawyers. Judges cited hallucination law 34
How could that many examples of fake cases produced by AIs and passed on by lawyers to support cases exist?
Remember, these are the ones that we know about. I wonder how many have been missed. times.
How much do these hallucination events undermine law practice? What needs to change?
I’ll let Lila weigh in.
AI, Lawyers and the Law
Every few months, we see new headlines predicting that AI is about to replace lawyers. These make for dramatic reading. Could AI end the need for lawyers? It certainly can if lawyers only confined themselves to drafting legal documents, summarizing case law, or searching for results in Westlaw and LexisNexis.
AI can indeed draft, summarize, and search through masses of documents faster than any lawyer. The practice of law, however, involves judgment, trust, and accountability, and that is something that AI cannot supply.
What is far more likely to change in legal practice is the way work gets done. What tasks will human lawyers spend their days on, and what will be the tasks given to AI to do?
Routine Tasks Are Being Automated First
The parts of legal work where AI fits are repetitive jobs: document review, contract comparison, legal research, and first drafts of standard forms. These jobs can be handled in a fraction of the time using AI, which can scan thousands of pages to find relevant clauses and flag inconsistencies that a human reviewer might miss.
That does not remove the need for a lawyer. Instead, it removes hours of tedious work, freeing attorneys to concentrate on analysis and strategy. Junior lawyers, who traditionally learned by doing this kind of grunt work, will see their training evolve, but the need for people who understand the law remains.
Note: there is no reference here to gathering case evidence for citing in a court of law.
Accountability Cannot Be Outsourced
Law is not simply a matter of retrieving the right precedent or rule. Lawyers weigh competing arguments, anticipate how a judge or opposing counsel might react, and advise clients on risk when the answer is uncertain.
While AI can produce confident-sounding text, it could be just plain wrong. That’s because AIs can hallucinate, citing cases that are not real. Why does this happen?
AI Large Language Models (LLMs) train on vast datasets and predict word sequences based on statistical likelihood. They don’t know truth from fiction, just patterns in the training data.
That’s why a lawyer who uses AI to search for court rulings to support an existing case needs to verify the content the AI feeds to them. The lawyer needs to own the case and supporting material, and answer to clients and regulators.
When a lawyer chooses to use AI to help in a case, they need to confirm that what they are being fed is real. Attorneys using AIs that produce fake case law are increasingly coming up in news stories.
There is no court on Earth where AIs can hold a license to practice the law, owe fiduciary duties, or be held accountable for mistakes. AI’s primary use in law should be restricted to assisting at what it does best: rapidly assessing and assembling content for human lawyers to review.
Clients Want and Need the Human Relationship
Many legal matters are deeply personal. People facing a divorce, a lawsuit, or the loss of a family member are rarely looking for a document generator alone. They want someone to listen, explain their options in plain language, and guide them through stressful decisions.
Estate planning is a good example, since families often need help navigating sensitive questions about money, relationships, and legacy.
A trust lawyer does much more than draft paperwork. They help clients think through goals, anticipate family dynamics, and choose structures that fit their circumstances. Empathy, discretion, and persuasion remain distinctly human strengths, and are often what clients value most.
New Skills and New Roles Are Emerging
As AI tools spread, the skills that set lawyers apart will shift. Attorneys who understand how to prompt, evaluate, and supervise the use of AI will work faster and offer more value.
Firms are already hiring legal technologists, data specialists, and process designers to help integrate new tools safely.
Questions about privacy, confidentiality, bias, and intellectual property also create fresh areas of legal work, since businesses need advice on how to use AI responsibly. Rather than shrinking the profession, technology may expand it by creating specialties that did not exist a decade ago.
Access to Legal Help May Improve
There is also an upside for the public. Many people cannot afford legal services or do not know where to begin. AI-assisted tools can lower costs, speed up routine matters, and help individuals understand basic rights and procedures.
Lawyers who adopt these tools may be able to serve more clients at lower prices, or offer flat-fee services that were previously impractical.
The risk is that poorly built tools could mislead users, which is why oversight and professional standards matter. Used wisely, AI could narrow the gap between those who can access legal help and those who cannot.
Law Practices Disrupted
The increasing use of LLMs and AI agents is changing law practice. It’s changing the human mix from law clerks to junior lawyers.
Understanding how to use AI tools while continuing to provide a client-centred focus going forward will be a balancing act. The future of legal practice, therefore, is looking less like a world without lawyers and more like a world where lawyers will use AI to produce better and more timely results for clients.

