Authored by Karla Rossouw
From doctor to therapist and now lawyer. When people find themselves facing a legal problem, one of the first things they often want is an answer. What does this letter mean? Can I be sued? Have I broken the law? What should I say? Nowadays, the first place people are turning to is an AI chatbot. It’s quick, available around the clock, and most importantly, it doesn’t feel judgemental.
That last part is more important than it might seem. People will often tell AI things that they would be reluctant to tell another person. They can explain exactly what happened, upload documents, name the people involved and ask questions about what could happen next. The conversation feels private and almost like speaking to a confidential adviser, but the important thing users need to understand is that AI isn’t your lawyer, and a conversation with an AI chatbot doesn’t automatically promise confidentiality.
The Problem Starts When AI Feels Like A Lawyer
Today’s chatbots no longer feel like traditional software and legal technology is changing how people interact with the law. You can explain a complicated situation in ordinary language, receive an answer that appears to understand what you’re dealing with, provide more context and ask the system to reconsider its response. Sounds pretty similar to a conversation with another person, right?
The Chancellor of the High Court in England and Wales described this experience as an “eerie sense” of communicating with a person. In a speech on legal professional privilege, he demonstrated this point by asking ChatGPT whether he could sue a builder who had broken his kitchen window. The chatbot responded with an explanation covering negligence, contract, remedies, and potential court action, before even requesting additional information. For someone dealing with a stressful legal problem, that can look extremely similar to a conversation with a lawyer.
What users tend to forget is that professional privilege exists to protect certain confidential information between a client and their lawyer. An AI chatbot doesn’t automatically create that relationship. Gowling WLG supported this argument in an analysis of AI and privilege by explaining that chatbots aren’t lawyers and that simply receiving an answer that looks like legal advice doesn’t create legal advice privilege.
More from Artificial Intelligence
- Salmon Introduces Execution Verification Infrastructure (EVI) For Securing AI Agents And Autonomous Systems
- 97% Off Claude And GPT: Why Stolen AI Subscriptions Are Flooding The Dark Web
- Most SMEs Are Using AI, But Very Few Are Seeing A Return
- Because AI Says So: What Happens When A Chatbot Becomes The Office Expert?
- Hotels Are Letting AI Agents Handle Bookings – What Could Go Wrong?
- When AI Copies Style In Seconds, Does Artistic History Become Obsolete?
- Can AI Help Us Better Understand Animal Behaviour?
- Your AI Agent Can Now Shop For You – Banks Say That Opens The Door To Scams
Your “Private” Chat May Not Be As Private As You Think
The problem arises when it comes to actual legal cases and privacy. For instance, a threatening letter arrives, and a user uploads it to ChatGPT and asks what it means. An employment dispute becomes tricky, and one of the parties gives AI access to the contract.
The Judiciary of England and Wales has already addressed this in its October 2025 guidance on artificial intelligence. Its guidance says users should treat anything they enter as though it could be published “to all the world,” which could potentially lead to a loss of legal privilege.
How Well Of A Lawyer Is Your AI Chatbot Actually?
If your AI chatbot went to law school, would it have passed the bar exam? In its 2026 judgment, the Upper Tribunal dealt with legal documents where AI had been used to find and cite legal cases. But when citations had to be verified, the mentioned cases were nowhere to be found. AI’s superpower is to hallucinate incorrect legal information and present it with 100% confidence.
AI chatbots do not provide answers from authoritative databases such as LexisNexis or Westlaw. Their responses are generated from the information available to the system and the prompt it receives, which means the result can be inaccurate, incomplete, misleading, or out of date. It could even go so far as producing fictitious cases, citations and other legal material that does not exist.
The Judiciary has mentioned that this is a particular concern for people representing themselves, who may not have the legal knowledge needed to realise when AI’s legal advice has gone off track.
Before You Confess Everything To The AI Chatbot
AI has made legal information much more accessible, but accessibility shouldn’t be confused with legal protection. If you’re dealing with a serious dispute, investigation, employment issue or potential criminal allegation, think carefully before uploading documents or explaining the circumstances to a public AI chatbot.
There is nothing wrong with using AI as a starting point. It can help explain unfamiliar legal terms or what to ask your lawyer, but when the answer could affect your case, your money, your employment or your freedom, that is where a qualified lawyer should take over. AI can help you understand the situation, but it should not be making decisions for you or in any way act as your lawyer.
Lawyers are able to give advice within a professional relationship that has legal rules around confidentiality and legal professional privilege which an AI chatbot cannot do. So, the next time you open a chatbot because you have a legal problem, pause before you spill all the details. The question isn’t just whether AI can give you sound legal advice; it’s whether you’d want the judge to read the entire chat as well.
