Not long ago, much of the legal profession’s attention turned to a series of cases in which lawyers were sanctioned for filing briefs that cited cases that did not exist. The culprit, we were told, was generative AI. The term “hallucination” entered the legal lexicon, and the cautionary tales came flooding in.

But there is a quieter, far more common problem that receives much less attention. It is the complaint that gets filed with allegations that have no factual basis, not because a machine made them up, but because the lawyer never took the time to find out whether they were true.

In my work handling legal malpractice claims, I see this pattern with some regularity. A complaint is filed. The allegations are serious. Sometimes they are career-threatening. And yet, when you begin to examine the underlying facts, it becomes clear that there was little to no investigation before those allegations were put on paper. The lawyer appears to have taken the client’s version of events at face value and run with it.

That is not just a strategic misstep. It is an ethical problem.

Rule 3.1 of the North Carolina Rules of Professional Conduct requires that a lawyer have a basis “in law and fact” for any claim asserted. That does not mean the lawyer has to prove the case before filing it. It does mean that the lawyer must have a good-faith belief that the facts exist or are likely to exist after reasonable investigation. Filing first and figuring it out later is not the standard.

Rule 3.3 reinforces the point. A lawyer may not make false statements of material fact to a tribunal. That obligation does not begin at trial. It begins the moment a pleading is filed. If you are putting factual allegations before a court, you are representing that you have a basis for making them.

So what does that look like in practice?

At a minimum, it requires more than listening carefully to your client and drafting artful allegations. It requires some level of independent judgment and verification. That might include reviewing key documents, obtaining and examining records, speaking with available witnesses, or at least testing the internal consistency of the client’s account. It also requires asking hard questions, especially when the allegations are serious or potentially damaging.

There is, of course, a line. Lawyers are not required to prove their case before filing suit, and some facts will only become available through discovery. But there is a meaningful difference between pleading facts that you reasonably expect discovery will support and pleading facts that you have done nothing to investigate.

This issue is not limited to factual allegations. It extends to legal claims as well. A recurring example is the inclusion of claims under North Carolina’s Unfair and Deceptive Trade Practices Act, Chapter 75, in lawsuits against lawyers. The statute expressly excludes “learned professionals” from its scope when they are providing professional services. While plaintiff’s lawyers often try to plead around the exception, these claims are often asserted without any basis.

That raises the same question under Rule 3.1. What is the legal basis for the claim? If the law clearly forecloses the theory, including it in the complaint is not creative lawyering. It is, at best, careless. At worst, it is sanctionable.

It is easy to treat the recent AI sanctions cases as something new and different. But at their core, they are not. They are reminders of an old principle that lawyers are responsible for what they file.

Whether the source is a chatbot or a client sitting across the desk, the obligation is the same. Before you put an allegation into a pleading, you have to do enough work to have a good-faith basis to believe it is true. Before you assert a claim, you have to know that the law supports it or that you are making a legitimate argument for changing it.

The real risk is not that technology will lead us astray. It is that we will use it, or our clients, as a substitute for our own judgment. And that is something no rule can fix for us.