Recently, there have been reports of lawyers submitting documents thinking, "It was written by AI, so it should be fine," only to discover that the case law did not even exist.

In short, they are getting caught for presenting fake case law documents to the court.

This is not just a few cases; it is happening in Korea, the U.S., and globally, leading to expressions like "it's a mess" in the legal community.

The core issue is that generative AI writes so convincingly.

The problem arises when one cannot distinguish whether the information mixed in is real or fake, which can lead to serious consequences. This is especially dangerous in the case of legal precedents.

These are not just ordinary documents; they are materials used as grounds for court judgments. Yet, there have been multiple instances where these were submitted without verification.

In California, actual disciplinary actions have already been taken. Several lawyers submitted documents created by AI that included non-existent case law and faced lawsuits or disciplinary measures. These lawyers were active in areas like LA, Beverly Hills, and Arizona, and the common factor was their use of AI without proper verification.

One case illustrates this more realistically. In a trademark lawsuit, a lawyer submitted non-existent case law as evidence.

This was not just a simple mistake; when the court checked, it turned out that such case law did not exist. In another case, a lawyer admitted to only verifying part of the information and submitting it without reviewing the whole.

Honestly, this is closer to "the AI was wrong" than "the person did not do their job."

There have already been cases where lawyers faced disciplinary actions for including fake case law in class action documents, resulting in suspension for wasting the court's time. The important point here is that the court takes this very seriously.

It is not a situation that can be brushed off as a simple mistake. They view it as an act that undermines the trust in the judicial system.

Why is this happening? Because AI is too convenient.

In the past, finding case law required spending time searching databases, reading each one, and organizing the information. Now, you can input a few lines and get a convincing document immediately. The problem lies in that "convincing" aspect. It creates "plausible-looking fakes" rather than actual facts.

This phenomenon is referred to in the industry as "hallucination." It is when AI generates something that is not true as if it were fact.

What's more frightening is that readers can easily be deceived. The sentence structure is natural, and the citation format is correct, so at first glance, it looks real. In a busy situation, if one thinks, "It should be fine," it can lead to immediate problems.

As a result, the atmosphere in the legal community is definitely changing.

Using AI itself is not a problem. In fact, it is widely used. However, the ultimate responsibility lies entirely with the lawyer.

"It was written by AI" is not an excuse. From the court's perspective, they only look at the results. If incorrect case law is presented, that is the lawyer's responsibility.

Is this issue limited to the legal field? Not at all. It applies across professional fields such as healthcare, finance, and accounting.

As we enter an era where AI writes for us, the role of humans has not diminished; rather, it has become clearer.

We have shifted to being "verifiers." If this is not done, it leads directly to problems.

In summary, the current situation can be boiled down to this:

"Those who can use AI well will not survive; those who can question AI will."