AI and Attorney-Client Privilege: What Should You Know?
Artificial intelligence tools are now part of everyday life. Millions of people use generative AI platforms to research medical conditions, draft emails, plan finances, and, increasingly, to research legal questions. When people face a personal injury claim, accident, or any legal matter, it is natural to turn to AI tools for quick answers and guidance.
What most people do not realize is that what they type into a public AI platform may not stay private. AI prompts are stored by the companies that operate these platforms. Depending on the platform’s privacy policy, those prompts can be accessed by the company, used to train future AI models, and, in some circumstances, disclosed to third parties, including government agencies. Courts across the United States are now actively deciding whether conversations a person has with an AI platform can be used as evidence in legal proceedings. The answer, in many cases, is yes.
What Are Some Recent Rulings on Attorney-Client Privilege and AI Use?
The most significant ruling to date on AI and attorney-client privilege came out of the Southern District of New York. In United States v. Heppner (2026), Judge Jed S. Rakoff ruled that documents a criminal defendant created using a publicly available AI platform were not protected by the attorney-client privilege or the work product doctrine.
Here is what happened: Bradley Heppner was indicted on federal fraud charges. After receiving a grand jury subpoena and retaining legal counsel, he used a public version of an AI platform to prepare materials outlining his defense strategy and potential legal arguments. When federal agents executed a search warrant at his home, they seized electronic devices containing those AI-generated documents along with his full prompt history. Heppner’s attorneys attempted to assert privilege over the materials. The court disagreed.
Judge Rakoff found that the AI-generated materials failed to meet the basic requirements for the attorney-client privilege on three grounds:
- No attorney-client relationship existed. The AI platform is not an attorney. Because Heppner was communicating with a third-party AI platform rather than with legal counsel, no attorney-client relationship was formed.
- No reasonable expectation of confidentiality. Heppner used a public version of the platform. That platform’s privacy policy allowed the company to collect user data, use it for model training, and disclose it to governmental regulatory authorities. The court found that under those terms, Heppner could not have had a reasonable expectation that his prompts would remain private.
- Not for the purpose of obtaining legal advice. The AI platform expressly disclaimed its ability to provide formal legal advice. Since Heppner was not acting at his attorney’s direction when he used the platform, the court found he was not seeking protected legal counsel.
The court also rejected the argument that the documents gained work product protection simply because Heppner later shared them with his attorneys. Privilege must exist at the time of the communication. Routing unprotected materials through an attorney after the fact does not make them privileged.
A separate 2026 ruling, Warner v. Gilbarco, Inc., reached a different result for a different reason. In that case, the court found that using AI tools did not automatically waive work product protection because the AI platform is a tool rather than a person, meaning disclosure to it is not the same as disclosure to an adversary. The key distinction between the two cases is whether an attorney directed the client’s use of the AI tool.
These rulings are setting real precedent. The extent to which defendants argue for protection over AI-generated materials will depend heavily on whether they used a public or private AI platform, whether their attorney directed the AI use, and whether the platform’s privacy policy supported any reasonable expectation of confidentiality.
What Is Protected Under the Attorney-Client Privilege?
Attorney-client privilege is one of the most important protections in the legal system. It allows you to speak openly and honestly with your attorney without fear that what you say will be used against you. Understanding what is and is not covered can help you protect your case.
To qualify for the attorney-client privilege, a communication must meet all three of the following conditions: it must be between a client and an attorney, it must be kept confidential, and it must be made for the purpose of obtaining or providing legal advice.
Information and communications generally protected by the attorney-client privilege include:
- Conversations with your attorney: about the facts of your case, your injuries, your concerns, and your legal options
- Written communications: including emails, letters, and messages exchanged directly with your attorney or their staff
- Documents you prepare at your attorney’s direction: to assist in building your legal strategy
- Information shared within the legal team: including paralegals and support staff who work on your case under an attorney’s supervision
- Legal strategies and case theories: developed through the attorney-client relationship
- Confidential communications: that reflect defense counsel’s strategy and mental impressions about your case
What is not protected includes conversations with friends or family about your case, information you post on social media, statements you make to insurance companies, and, as the Heppner ruling demonstrates, prompts you enter into a public third-party AI platform on your own initiative.
How Do I Protect Myself and My Case During the Legal Process?
If you are involved in a personal injury claim or any other legal matter, your behavior outside of your attorney’s office can directly affect the strength of your case. Here are practical steps to protect yourself:
Avoid using public AI tools to research or plan your legal strategy. Typing questions about your case into a generative AI platform, even casually, creates a record that may not be protected. The AI platform’s privacy policy, not your intentions, determines how that data is handled.
Do not discuss your case on social media. Posts, comments, messages, and even deleted content can be discovered and used as evidence.
Communicate about your case only through your attorney. Information shared directly with your legal counsel is protected. Information shared with anyone else, including a third-party AI platform, generally is not.
Ask your attorney before using any AI tools in connection with your case. There is an important distinction in the law. When a client uses AI tools at their attorney’s direction, courts have indicated the resulting materials may stand on stronger footing. When a client acts independently, that protection disappears.
Be mindful of privacy policies. If your attorney recommends using a specific secure platform or enterprise AI tool, understand that consumer-grade, publicly available AI platforms are not built for legal confidentiality. Their privacy policies often allow data to be used for training and shared with third parties.
Talk openly and honestly with your attorney. The attorney-client relationship is built on trust. Your attorney cannot protect you without complete and accurate information. The legal protections that exist are designed to make those conversations safe.
Why Is Fran Haasch Law Group Accident & Injury Lawyers Writing About AI Use?
Fran Haasch Law Group Accident & Injury Lawyers represents people who have been hurt through no fault of their own. That means being straightforward with clients about every factor that can affect the outcome of their case, including factors they may not have considered.
AI tools have changed how people research and prepare for legal proceedings. Most clients are not lawyers. They do not know that typing their injuries, accident details, and concerns into a public AI platform could give the other side a window into their legal strategy. They do not know that a grand jury subpoena or a search warrant can reach those AI-generated documents. And they do not know that a court may find no attorney-client privilege applies, as happened in United States v. Heppner.
Fran Haasch Law Group has built its reputation on decades of experience in personal injury law and a deep commitment to protecting the rights of every client. That commitment means more than skilled courtroom representation. It means making sure clients understand the full picture before they make decisions that could affect their case.
Protecting the trusting human relationship between attorney and client is a cornerstone of personal injury representation. Every communication that stays within the protected attorney-client relationship is a communication that cannot be weaponized against you. Every prompt entered into a public AI platform on your own is a risk that did not have to be taken.
If you have been injured and have questions about your case, speak directly with an attorney. That conversation is protected. The one you have with an AI platform may not be.
