Individuals, business owners, and executives are increasingly using ChatGPT, Claude, and other AI platforms as a stand-in for a lawyer: evaluating disputes, drafting contracts, planning negotiations, and even discussing potential criminal exposure. But AI is not your attorney, and a conversation with an AI platform does not automatically carry attorney-client privilege.
Why this matters
Attorney-client privilege generally protects confidential communications between a lawyer and client made to obtain or give legal advice. But confidential and privileged are not the same thing. When you voluntarily share sensitive legal information with a third-party AI platform, you may lose protections you otherwise would have had. Depending on the platform’s terms, your settings, and applicable law, those conversations may be reachable through discovery or lawful government process.
Recent decisions show this is already happening
United States v. Heppner (S.D.N.Y., Feb. 17, 2026)
A defendant facing federal securities and wire-fraud charges used Claude to prepare roughly 31 documents analyzing his defense, then shared them with his attorneys and claimed privilege. Judge Rakoff ruled the documents were not protected by attorney-client privilege or work product. Claude is not an attorney, the defendant created the documents on his own rather than at his lawyers’ direction, and sharing with a third-party platform undermined confidentiality.
Takeaway: even information that began with your lawyer can lose protection once you enter it into an outside AI platform.
Fortis Advisors v. Krafton (Del. Ch., Mar. 16, 2026)
In a dispute over a potential $250 million earnout, Krafton’s CEO used ChatGPT to develop a takeover strategy, including ways to avoid the earnout. Those conversations became evidence of the company’s intentions and decision-making, and the court ultimately restored the terminated executives to their positions.
Takeaway: AI chats can give opposing counsel a detailed record of your motives and knowledge.
Warner v. Gilbarco (E.D. Mich., Feb. 10, 2026)
A self-represented plaintiff used a paid ChatGPT account to prepare litigation materials. The court denied the other side’s request for those conversations, finding them protected work product.
Takeaway: the law is unsettled, and courts look closely at the purpose and circumstances of AI use.
Morgan v. V2X (D. Colo., Mar. 30, 2026)
The court recognized that AI-assisted litigation preparation can qualify for work-product protection, but required the plaintiff to identify the platform used and imposed safeguards on putting confidential materials into platforms lacking adequate contractual privacy and security protections.
Takeaway: uploading contracts, discovery, or business records to AI can create new discovery and confidentiality problems.
What you should do
- Treat anything you type into an AI platform as something an opposing party, regulator, or judge could one day read.
- Do not enter privileged communications, legal strategy, or confidential business information into public AI tools.
- Talk to your attorney before using AI on a dispute, investigation, transaction, or contract.
Questions about your situation? Contact Staller Law to discuss how to use AI safely without putting your legal position at risk.
This alert is for general informational purposes only and does not constitute legal advice or create an attorney-client relationship.