--- title: "Using AI as Your Injury Lawyer: 4 Risks to Know | Sam & Ash" date_published: "2026-09-21" date_modified: "2026-09-21" permalink: "https://samandashlaw.com/resources/using-ai-as-your-lawyer-after-an-injury-heres-what-it-can-cost-you/" taxonomies: Categories: - "Blog" Tags: - "personal injury" seo: title: "Using AI as Your Injury Lawyer: 4 Risks to Know | Sam & Ash" description: "A chatbot can explain a legal term. It can't protect your claim, it can't be held accountable, and it can't keep your secrets." canonical_url: "https://samandashlaw.com/resources/using-ai-as-your-lawyer-after-an-injury-heres-what-it-can-cost-you/" --- A chatbot can explain a legal term. It can't protect your claim, it can't be held accountable, and it can't keep your secrets. Quick answer: AI chatbots like ChatGPT, Claude, Grok, and Gemini are not lawyers and can't act as one. They can hand you wrong legal information with complete confidence, lose track of details you gave them (including deadlines), and your conversations with them are generally not protected by attorney-client privilege, which means they may be discoverable in a lawsuit. If a chatbot gets your injury claim wrong, there's no malpractice claim and no one who answers for it. You can use AI to learn vocabulary. Don't use it to run your case. Something new keeps landing on our desks. Injured people are turning to attorneys after weeks, sometimes months, of letting a chatbot guide their claim. They've asked AI what their case is worth, whether to give the adjuster a recorded statement, and how long they have to file. Some of what came back was fine. Some of it was dangerously wrong. Recently, our attorneys reviewed a client's conversation with an AI chatbot. Early in the chat, he told the bot about a critical deadline in his case. Later in that same conversation, the bot advised him as if that deadline didn't exist. No flag. No warning. It just forgot. That's the whole problem in one sentence: when AI makes a mistake on your injury claim, you're the only one who pays for it. ## Why Insurance Companies Don't Mind If You Use AI Instead of a Lawyer Insurers have spent years building their own AI. In a 2021 survey run under the authority of regulators in nine states, including Nevada, [88% of large auto insurers said they use, plan to use, or plan to explore AI and machine learning](https://content.naic.org/sites/default/files/inline-files/BDAIWG%202021%20PPA%20Survey%20Memo%20120822.pdf), and 70% reported using it in claims. The company on the other side of your claim also has data analysts, claims software, and a legal department. If all you have is a chatbot, that's the matchup they want. Insurers profit when injured people go it alone, and someone typing questions into an app at midnight is still going it alone. The chatbot won't call the adjuster, won't track down surveillance footage, won't hire an accident reconstructionist, and won't push back or take anyone to trial when a lowball offer shows up. ## Pitfall #1: AI Invents Law and Sounds Sure About It AI "hallucinations" are false statements delivered with complete confidence. In law, they're common. A [Stanford RegLab and HAI study](https://hai.stanford.edu/news/hallucinating-law-legal-mistakes-large-language-models-are-pervasive) found that the leading AI models of the time hallucinated 69% to 88% of the time when asked specific, verifiable questions about federal court cases, and they often failed to correct a user's wrong legal assumptions. Even paid research tools built for lawyers [gave incorrect information more than 17% of the time, about one in six queries](https://hai.stanford.edu/news/ai-trial-legal-models-hallucinate-1-out-6-or-more-benchmarking-queries), in a follow-up Stanford study. This isn't theory. In 2023, a federal judge sanctioned two New York lawyers and their firm [$5,000 after they filed a brief citing court decisions ChatGPT made up](https://gc.ai/blog/ai-hallucination-legal-cases) (Mata v. Avianca). Since then, researcher Damien Charlotin's [AI Hallucination Cases database](https://www.damiencharlotin.com/hallucinations/) has logged more than 2,000 court decisions worldwide involving AI-fabricated material, and [most of the cases tracked didn't involve a practicing lawyer](https://legalaispace.com/blog/ai-hallucination-cases-law-firms-2026). If trained lawyers get burned, an injury victim without legal training has almost no way to catch the error before it does damage. ## Pitfall #2: AI Forgets What You Told It Chatbots don't remember the way people do. They process your conversation as one long block of text, and they don't weigh all of it equally. A [Stanford-led study published in 2024](https://aclanthology.org/2024.tacl-1.9) found that language models perform best when key information sits at the beginning or end of a long input, and significantly worse when it's buried in the middle. Now picture your claim. You mention your accident date on Monday. You spend the week asking about medical bills, the adjuster's emails, and lost wages. By Friday, the detail that matters most, your deadline, is sitting in the middle of a very long chat. That's what happened to our client. In Nevada, you generally have [two years from the date of injury to file a personal injury lawsuit](https://www.leg.state.nv.us/nrs/nrs-011.html) under NRS 11.190(4)(e). California's general limit is also [two years](https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=CCP§ionNum=335.1), but if a government agency is involved, you may have only [six months to file a claim](https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=GOV§ionNum=911.2). Miss the deadline, and a strong case can be thrown out no matter how badly you were hurt. ## Pitfall #3: Your AI Chats Aren't Privileged and May Be Discoverable What you tell your lawyer is protected by attorney-client privilege. What you tell a chatbot isn't. In February 2026, in United States v. Heppner, federal Judge Jed Rakoff ruled that [31 documents a defendant generated with an AI chatbot weren't privileged](https://harvardlawreview.org/blog/2026/03/united-states-v-heppner/). The court gave three reasons: [the AI isn't a lawyer, the platform's privacy policy undercut any expectation of confidentiality, and the defendant used it on his own rather than at his lawyer's direction](https://www.orrick.com/en/Insights/2026/03/Court-Rules-AI-Conversations-Are-Not-Privileged-What-United-States-v-Heppner-Means-for-You). Handing the chats to his lawyers afterward [didn't make them privileged after the fact](https://www.frierlevitt.com/articles/ai-attorney-client-privilege-united-states-v-heppner/). The people who build these tools say the same thing. OpenAI CEO Sam Altman [acknowledged](https://futurism.com/chatgpt-legal-questions-court) there's no legal confidentiality for ChatGPT conversations and that, in a lawsuit, "we could be required to produce that." A federal court has already ordered OpenAI to [turn over 20 million de-identified ChatGPT logs](https://decrypt.co/351000/openai-ordered-20m-chatgpt-logs-nyt-copyright-case) in copyright litigation. The law is still developing. The same week as Heppner, a federal court in Michigan [refused to force a self-represented plaintiff to turn over AI-assisted materials](https://abovethelaw.com/?p=1179748) on work-product grounds. But you don't want your injury claim to be the test case. Here's why it matters: in an injury lawsuit, the insurer's lawyers can request documents about your accident, your injuries, and your damages. If you typed "my back felt fine until a week later" or "I was going a little fast" into a chatbot, that's a written statement in your own words. Defense lawyers are trained to find exactly those lines. ## Pitfall #4: When AI Gets It Wrong, You Have Zero Recourse When a lawyer blows a deadline or gives negligent advice, you have options. A lawyer owes you loyalty and confidentiality, answers to the State Bar, and can be held liable for malpractice. In Nevada, [with narrow exceptions, only active members of the State Bar can practice law](https://www.leg.state.nv.us/nrs/nrs-007.html) (NRS 7.285). A chatbot owes you none of that. The FTC went after DoNotPay, which marketed itself as "the world's first robot lawyer," after finding the company [never tested whether its AI performed like a human lawyer and never hired attorneys to check its accuracy](https://www.ftc.gov/news-events/news/press-releases/2025/02/ftc-finalizes-order-donotpay-prohibits-deceptive-ai-lawyer-claims-imposes-monetary-relief-requires). It had to pay $193,000 and stop making those claims. And OpenAI's own [usage policies](https://openai.com/policies/usage-policies/) bar using its services for tailored legal advice without a licensed professional involved. Read that again: the company that makes the chatbot says the chatbot shouldn't be your lawyer. ## What AI Can and Can't Do for Your Injury Claim Reasonable useLeave it to a lawyerExplaining a general term like "comparative negligence"Deciding what your claim is worthHelping you write questions for a free consultationTracking your filing deadlinesBackground on how injury claims generally workDeciding whether to give a recorded statementResponding to the adjuster or accepting an offer Even for reasonable uses, keep the facts of your accident out of the chat. ## Steps to Take If You've Been Using AI on Your Injury Claim Stop adding case details. Don't describe your injuries, who was at fault, or your conversations with the adjuster to a chatbot. Don't delete your chats. Once a claim or lawsuit is on the horizon, destroying relevant records can create serious legal problems. Leave them alone and tell your attorney they exist. Treat every AI number and deadline as unverified. Settlement values, filing dates, and legal rules from a chatbot mean nothing until a licensed attorney confirms them. Don't give a recorded statement or sign anything based on AI advice. Gather the real evidence: medical records, photos, the police or incident report, witness contact information, and every letter or email from the insurer. Talk to a licensed injury attorney now, while the clock is still on your side. ## Frequently Asked Questions ### Can I use ChatGPT as my lawyer after a car accident? No. ChatGPT and similar tools aren't licensed to practice law, can give confidently wrong answers, and don't owe you any legal duties. OpenAI's own policies say its services shouldn't provide tailored legal advice without a licensed professional involved. ### Are my AI chatbot conversations confidential? Generally, no. A federal court ruled in 2026 that a person's chats with a consumer AI tool weren't protected by attorney-client privilege, and AI companies can be compelled to produce user data in litigation. ### Can the insurance company get my AI chat history? Potentially. If your case goes to litigation, the defense can request documents related to your accident and injuries, and courts are increasingly treating AI chats like any other written record. ### Should I delete my AI conversations about my accident? No. Deleting records that may be relevant to a claim can hurt you. Stop adding to them and tell your attorney they exist. ### What happens if an AI chatbot gives me bad legal advice? In practice, very little. There's no attorney-client relationship, no malpractice claim against a licensed lawyer, and no State Bar to hold anyone accountable. The consequences land on you. ### How long do I have to file a personal injury claim in Nevada? Generally two years from the date of injury under NRS 11.190(4)(e). Some claims, including those involving government entities, have different rules, so confirm your deadline with an attorney. ## You Deserve a Real Fight, Not a Guess Your injury claim is too important to hand to software that can forget your deadline, invent the law, and leave a paper trail the insurer can ask for. At Sam & Ash Injury Law, a licensed attorney tracks your deadlines, protects your privileged communications, and stands between you and the insurance company, whether your case is worth $15,000 or $1.5 million. Consultations are free, available 24/7, and you owe nothing unless we win. Call Nevada at 702-857-6566, California at 949-694-3120, or our general line at 1-877-658-9951. The first conversation is free, and so is the honesty. Ask a chatbot for a recipe. Call us for your case. Sources: [Harvard Law Review: United States v. Heppner](https://harvardlawreview.org/blog/2026/03/united-states-v-heppner/) [Orrick: Court Rules AI Conversations Are Not Privileged](https://www.orrick.com/en/Insights/2026/03/Court-Rules-AI-Conversations-Are-Not-Privileged-What-United-States-v-Heppner-Means-for-You) [Frier Levitt: AI Conversations and Attorney-Client Privilege](https://www.frierlevitt.com/articles/ai-attorney-client-privilege-united-states-v-heppner/) [Above the Law: Warner v. Gilbarco analysis](https://abovethelaw.com/?p=1179748) [Futurism: Sam Altman on ChatGPT and legal confidentiality](https://futurism.com/chatgpt-legal-questions-court) [Decrypt: OpenAI ordered to produce 20M ChatGPT logs](https://decrypt.co/351000/openai-ordered-20m-chatgpt-logs-nyt-copyright-case) [Stanford HAI: Hallucinating Law](https://hai.stanford.edu/news/hallucinating-law-legal-mistakes-large-language-models-are-pervasive) [Stanford HAI: AI on Trial, legal models hallucinate in 1 of 6 queries](https://hai.stanford.edu/news/ai-trial-legal-models-hallucinate-1-out-6-or-more-benchmarking-queries) [Liu et al., "Lost in the Middle," TACL 2024](https://aclanthology.org/2024.tacl-1.9) [Damien Charlotin: AI Hallucination Cases Database](https://www.damiencharlotin.com/hallucinations/) [GC AI: AI Hallucination Sanctions Tracker (Mata v. Avianca)](https://gc.ai/blog/ai-hallucination-legal-cases) [LegalAI Space: 1,600+ AI Hallucination Cases](https://legalaispace.com/blog/ai-hallucination-cases-law-firms-2026) [FTC: Final order against DoNotPay](https://www.ftc.gov/news-events/news/press-releases/2025/02/ftc-finalizes-order-donotpay-prohibits-deceptive-ai-lawyer-claims-imposes-monetary-relief-requires) [OpenAI: Usage Policies](https://openai.com/policies/usage-policies/) [NAIC: 2021 Private Passenger Auto AI/ML Survey memo](https://content.naic.org/sites/default/files/inline-files/BDAIWG%202021%20PPA%20Survey%20Memo%20120822.pdf) [Nevada Revised Statutes, Chapter 11 (NRS 11.190)](https://www.leg.state.nv.us/nrs/nrs-011.html) [Nevada Revised Statutes, Chapter 7 (NRS 7.285)](https://www.leg.state.nv.us/nrs/nrs-007.html) [California Code of Civil Procedure 335.1](https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=CCP§ionNum=335.1) [California Government Code 911.2](https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=GOV§ionNum=911.2)