Chatting with AI Could Be Evidence, US Court Ruling Sparks Concerns

Jakarta - In an era where people confide in AI faster than to lawyers, a serious warning has emerged from the United States legal community: conversations with chatbots like ChatGPT and Claude can be used as evidence in court.

The warning comes after a recent federal court ruling in New York that sparked widespread concern among law firms. More than a dozen major firms have now issued official admonitions to clients to be careful about using AI when facing legal proceedings.

The Case That Became the Turning Point

The ruling came from Judge Jed Rakoff, who ordered a defendant to hand over documents created using AI as part of his defense preparations.

The case involves a former financial company executive who used Claude to compile materials related to his case, then shared them with his legal team. The legal team argued that the documents should be protected by the attorney-client privilege principle. However, the court took a different view.

In his ruling, the judge stated that there was no—and could not be—a lawyer-client relationship between the user and the AI platform. In fact, it was stated that users had no expectation of privacy over the input they entered into the system.

Why is it not protected?

In the US legal system, communications between clients and lawyers are protected by the attorney-client privilege. However, this protection can be lost if information is shared with third parties.

Here's the problem: AI is considered a third party, not a lawyer. That is, when someone types in the details of his case into a chatbot, he is legally considered to have opened the information outside the scope of protection.

Both OpenAI and Anthropic in their terms of service mention that user data can be shared with third parties under certain conditions.

The impact is not trivial. Information entered into AI - from legal strategies to case details - can potentially be requested by prosecutors or opposing parties in court.

For law firms, this is a loud alarm. Many are now advising clients not to use AI in sensitive contexts, especially when involved in litigation.

The New Era of Digital Risk

This phenomenon marks a major change in the way technology interacts with the law. AI does make many things easier - from drafting documents to quick analysis - but it also carries new risks that were previously unthinkable.

In this context, the old principle still applies: whatever you share with a third party can be evidence. The difference now is that the "third party" is not a human ... but an algorithm.

Follow VOI Whatsapp Channel