If you’re going through a divorce, custody dispute, or other family law matter in New Jersey, it’s tempting to turn to Ai like ChatGPT, Claude or a similar AI chatbot for quick answers thinking perhaps you’re saving money by not talking to your lawyer, or helping organize your thoughts and improve your outcomes. You might ask it to explain a legal term, help you organize your finances, or even draft a parenting plan. It feels private, fast, free and reliable. But none of those things may be true.
But recent court rulings make clear that what you type into a public AI chatbot may not stay private — and it could end up as evidence used against you.
AI Â Conversations May Not Be Confidential
Many people assume a conversation with an AI chatbot is like a conversation with a friend or a search engine — something that disappears into the void. It isn’t. When you use a consumer AI tool, your inputs are typically stored on the company’s servers, and depending on the platform’s terms of service that data may be used, retained, or accessed in ways you don’t control. It could even be supboen’d and used in court against you.
Courts have started to weigh in on exactly this issue. In one recent federal case, a court found that a user had no reasonable expectation of confidentiality in AI chat logs because the platform’s terms of service allowed the company to collect, use, and even disclose that information to third parties. Other federal rulings have gone further, requiring litigants to actually turn over their AI chat histories during discovery.
For family law litigants, this is a serious problem. Attorney-client privilege protects confidential communications between you and your lawyer. It does not protect what you type into a chatbot — because a chatbot is not your attorney, and OpenAI’s servers are not a law office.
What This Means for Your NJ Family Case
New Jersey’s discovery rules, like most states’, are broad. Parties are generally required to produce relevant electronically stored information, and that can include chat logs, drafts, and other AI-generated content. If your spouse’s attorney learns — or even suspects — that you’ve been using AI to work through your case, they may request your full chat history as part of discovery.
Think about what that could expose:
– Custody strategy. If you’ve asked ChatGPT to help you brainstorm arguments about your co-parent’s fitness, or to help you frame incidents in a more favorable light, that reasoning could be turned over and used to challenge your credibility.
– Financial information. Uploading bank statements, income details, or asset lists to get help “organizing” them creates a record — one that could later be scrutinized for inconsistencies or omissions.
– Â Case-sensitive admissions. Â A seemingly innocent question like asking whether a certain purchase or communication could be “a problem” in your case can look very different in front of a judge than it did in the moment you typed it.
– Draft agreements or provisions. Â If you use AI to draft or revise settlement language, those drafts and the surrounding conversation may become discoverable, potentially undercutting your negotiating position.
AI Also Gets the Law Wrong – A LOT
Separate from the privacy issue, generative AI tools are known to fabricate case citations, misstate legal standards, and present confident-sounding but incorrect information as fact.  Even lawyers are getting in trouble and facing sanctions and ethics complaints when they use AI without fact checking and it turns out to be wrong. New Jersey courts and the New Jersey Supreme Court’s preliminary guidance on AI have made clear that this doesn’t just create ethical exposure for attorneys who fail to verify AI output — it can directly harm a self-represented party who files an AI-drafted document that misstates the law or relies on cases that don’t exist.
Family law issues like custody determinations, alimony, and equitable distribution in New Jersey depend on specific statutes and case law (like the best-interests-of-the-child factors under N.J.S.A. 9:2-4). A general-purpose chatbot has no reliable way to confirm it’s applying current, accurate New Jersey law to your specific facts — and if it’s wrong, you’re the one who bears the consequences in court. Don’t worry, AI will sleep just fine as you get embarrassed in court or worse.
What to Do Instead
– Don’t discuss your case with a public AI chatbot — not your strategy, your finances, your co-parent’s behavior, or anything you wouldn’t want a judge to read.
– Don’t upload documents related to your case (financial records, correspondence, draft agreements) to a consumer AI tool.
– Treat anything you’ve already typed into a chatbot as potentially discoverable, and talk to your attorney about it before it becomes an issue in litigation.
– Bring your questions to your attorney instead. What feels like a quick shortcut can create a permanent, producible record that undermines the very case you’re trying to protect.
– You CAN use AI to help clarify terminology your lawyer or the court uses but don’t rely on it alone.   Be highly skeptical of any answer it gives you and confirm.
– You can also use AI to look for resources for information, for example  prompting it to point you to primary sources such as where the custody statute is for you to review it in it’s original form
If you have questions about how AI use — yours or your co-parent’s — might affect your New Jersey family law matter, talk to your attorney before you type another word into that chat box. Free advice can be the most expensive advice of them all.