After being injured in an accident, the next logical step is to do some research to figure out your rights and options for recovery. Like most people, I search everything online. And using AI engines and chatbots is often where people end up, in part because sites drive traffic to those models even when people may not have meant to get a response from AI.
If a person is injured and has a case against the at-fault party, it makes sense to research what to anticipate in the legal process. However, a person’s search and chat history often contains private information they would not want to share.
Recent court rulings have found that a person’s AI search and conversation history could actually be admitted in court. This raises a serious risk that people’s innocuous research could be selectively used against them in court.
Is Your AI Conversation Truly Private?
While a person’s conversations with an attorney are kept strictly confidential under the doctrines of attorney-client privilege and the work product, courts have recently ruled that AI searches/chats conducted by the person who is represented are not protected under this doctrine.
The first case to examine where this issue arose, is in the context of a criminal fraud case, United States v. Heppner1. The defendant was charged with a money laundering scheme and had generated documents using Claude (Anthropic’s AI Model) that outlined legal defense strategies. He had later shared these documents with his attorneys, and some of the information he put into Claude he had learned from his attorney. The criminal prosecutors sought to use these documents against him at trial. The court noted that attorney-client privilege attaches to communications (1) between a client and their attorney, (2) made in confidence, and (3) for the purpose of obtaining or providing legal advice. The court found the AI chatbot records were not a communication between the client and their attorney, were not confidential, and that the communications were not for obtaining legal advice because the chatbot expressly disclaims that they are providing legal advice. The court also found the work-product doctrine didn’t apply because they were not prepared by the defendant’s counsel and they did not reflect the attorney’s litigation strategy. It was important that the defendant had conducted the searches without any input from his attorney, and the court noted that the result may have been different if that had been the case.
In a different setting, a court in the civil case Warner v. Gilbarco2 did not allow AI searches and records to be disclosed to the opposing party where the person was representing themselves in court (this is also called “appearing pro se”). Although the opposing party argued that the person had waived protection of this information by sharing the information with a third party (the AI platform), the court found that the protection would be waived only by sharing information in a manner that might be shared with an adverse party. A similar ruling was issued in a separate matter, Morgan v. V2X, Inc.3, which also involved a pro-se litigant. These rulings both centered around the fact the person was representing themselves in the court proceedings.
Overall, the recent rulings suggest that legal research, searches, and queries through AI platforms by lawyers, or people who choose to represent themselves, are still protected work product and cannot be disclosed to the other party.
AI Examples
The landscape is constantly evolving, and courts do not have clear rules. What we know (so far!) is that when using AI in the context of a court case, the safest bet is to NOT use AI to assist you with your legal case. AI is not a lawyer and cannot give legal advice, even if it doesn’t say so, as in the examples below:
Two Must-Know Tips when Using AI
- If you have an attorney and want to use AI to work on your case, consult with your attorney before doing so. The rules are constantly changing and your attorney can give you the most up-to-date information on how to make sure that you are able to answer the questions you want asked and are able to protect yourself from the defense attorneys trying to make you look bad.
- If you are going to use AI on your own, search wisely. Think about how a question might be taken out of context and used against you BEFORE you interact with an AI platform. For example, if you want to know what your case is worth, rather than asking AI, “How much money can I get from a car crash?” consider how you would feel if that were read aloud by a defense attorney to the jury that you are asking to fairly compensate you for your injuries.
Key Takeaways
The existing court rulings put injured people in a bind. Most people do not have access to a 24/7 attorney who can provide legal advice on any given subject (I certainly don’t). But be aware that when using Chat GPT, the conversation can be used against the injured person in court.
As we anticipate many more court cases addressing these issues, we will keep you posted. In the meantime, if you find yourself needing an immediate response from an attorney, call our office to schedule a free consultation. (206) 682-0300.
