ARTIFICIAL INTELLIGENCE BURNS A NEW YORK LAWYER

 

One of the first things a student learns in law school is how to research and write a brief.  This is a difficult skill to learn, and writing a good brief can take considerable time and expense.

When I entered law school, two electronic research aids had just appeared: Westlaw and Lexis.  These were electronic databases that saved lawyers considerable time and allowed all of us to write better briefs.

Electronic databases have become so prevalent that many law libraries no longer house "real" books.  the only resource for legal research has been, for the last 10-15 years, the electronic database online.

In the past two years, a major breakthrough has occurred in this area: the use of artificial intelligence in researching and writing legal briefs. Many lawyers have sought to cut their research and writing time significantly through the use of AI services.  One such resource is ChatGPT.

This service not only researches your issue, but it also writes your brief. That saves lawyers a TON of time.

The problem with the service is that it has not provided the level of accuracy that courts absolutely demand when being told why a decision should go one way or another.  Mr. Steven Schwartz of New York discovered this truth the hard way.  

When Mr. Schwartz submitted a brief to a New York court describing how other courts had agreed with his position in his case, he did not check the accuracy of the citations.  the AI source had made up fake cases that supported his position.  The opposing attorney, doing his due diligence to prepare to address why the court should consider his own authorities as outweighing the arguments of opposing counsel, discovered the authorities cited in support of Mr. Schwartz did not even exist.

That is a major ethical violation.  While a lawyer is duty-bound to argue on behalf of his client, he is also duty-bound to be absolutely honest with the court.  Failure to do so can result in paying the other party's attorney fees for having to respond to a fake argument, paying a fine for contempt of court, and in a proceeding very likely to find its way to the attorney's bar association, it could result in the loss of the attorney's license to practice law.

In his defense, Mr. Schwartz told the court he had no idea that an AI research firm would make up fake cases to support a position.  

In all candor, I had no idea that could happen either.  However, Mr. Schwartz took too many shortcuts.  He absolutely should have checked the existence of those cases cited and verified by page and line that his authorities cited actually existed, and that they were not misquoted or taken out of context.

Judge P. Kevin Castel has taken the matter under advisement and will issue a decision concerning what sanctions to impose at a later date.  

I expect the penalties will be severe.