The most consequential development is a February 2026 ruling from Manhattan federal court. Judge Jed Rakoff, calling it "a question of first impression nationwide," held that 31 documents a defendant created with a consumer AI chatbot were not protected by attorney-client privilege. The defendant had entered information from his defense lawyers into the tool and sent the outputs back to them, arguing that this made the material privileged. The court disagreed. The exact scope of the ruling's reach in state family courts is not yet settled. The Manhattan ruling and the Cordell & Cordell figures come from secondary reporting and law-firm commentary; the underlying court documents have not been reviewed.
For family lawyers, the practical risk is discovery. Illinois discovery rules allow broad document requests in contested divorces, and commentators expect courts to reject sweeping demands for "all AI activity." Targeted requests tied to finances, assets, or litigation strategy are more likely to succeed. Lawyers also report that clients who consult a chatbot first often arrive with unrealistic expectations. One attorney quoted by Axios called it "the WebMD effect on steroids." Others say better-informed clients can help.
Attorneys should ask new clients what tools they have used and advise them in writing that chatbot histories may be discoverable. Check every citation in documents clients draft, and be ready to move for sanctions when filings contain fabricated authority.