A federal judge in New York has ruled that material generated through a public AI chatbot is not protected by attorney-client privilege or the work-product doctrine, a decision that could shape how lawyers and clients use generative AI in litigation. According to reporting by legal publishers and commentary from law firms, the case arose after a defendant used an AI system to help develop defence theories and later sought to shield the resulting exchanges from disclosure. The court rejected that effort, finding that the communications were not confidential in the way the privilege rules require.

The ruling, issued by U.S. District Judge Jed S. Rakoff in the Southern District of New York, turns on a familiar principle rather than a novel rule for new technology: privilege depends on secrecy. Legal commentators have said the defendant had used a publicly accessible chatbot without lawyer involvement, meaning the information was shared with an outside platform rather than exchanged privately with counsel. On that basis, the court treated the AI system as a third party, which undermined any claim that the material stayed within the protected attorney-client relationship.

The decision is also notable because it extends beyond privilege to work-product protection. Akerman reported that the judge concluded the AI-related material was discoverable even after the defendant passed search results to counsel. Other legal analyses said the court’s reasoning emphasised that AI tools are not licensed legal advisers and therefore do not fit neatly within doctrines built around lawyer-client communications. The result is one of the first federal rulings to squarely address whether conversations with generative AI can be kept out of litigation, and it came down firmly on the side of disclosure.

For lawyers and in-house teams, the practical message is cautionary rather than prohibitive. The case does not suggest that AI cannot be used in legal work, but it does underline the risk of feeding sensitive facts, strategy or draft arguments into consumer-facing tools that may store, process or reuse inputs. As several legal firms have noted, public AI services often reserve rights over user data, making it harder to argue that a client had a reasonable expectation of privacy.

That leaves companies and individuals facing a simple but significant rule: if confidential information is shared with a third-party AI system, traditional privilege protections may be lost. The broader implication of the ruling is that courts are likely to apply established confidentiality standards to new technology, rather than carving out special treatment for artificial intelligence.

Source Reference Map

Inspired by headline at: [1]

Sources by paragraph:

Source: Noah Wire Services