I've recently become aware of a serious legal risk issue. Over a dozen top U.S. law firms are warning their clients that discussing legal matters with AI chatbots could be very problematic.

The origin of this issue is a ruling by Judge Jed Rakoff in New York in February. In a fraud case, the judge ruled that conversations between the defendant and AI chat tools like Claude, along with 31 AI-generated defense documents, could be used by the prosecution. What does this mean? It means these communications are no longer protected by the attorney-client privilege.

Once this ruling was made, law firms became uneasy. They began to urgently notify clients to avoid asking sensitive legal questions to AI chatbots. The reasoning is simple: what you discuss with AI might not have any legal protection, and once litigation begins, those chat records could be subpoenaed and used as evidence by the opposing counsel.

The current problem is that many people are not fully aware of this. AI chat tools are indeed convenient, but the legal risks are especially high. The attorney-client privilege is critically important in the U.S. legal system; it protects confidential communications between you and your lawyer. But if you bypass a lawyer and go directly to an AI chatbot, that protection no longer applies.

So, this is not just a technical issue but a challenge to the entire legal ecosystem adapting to the AI era. More similar rulings may emerge, and regulations will gradually become clearer. People with legal needs are best advised to consult professional lawyers directly and not rely too heavily on AI chat tools for sensitive matters.