The software ran on two websites maintained by Henry Ihejirika -- Ziinet.com and 700law.com -- which offered automated assistance for people going through bankruptcy proceedings.

Users would enter their personal information and other relevant details onto the website, and the AI would generate a complete set of bankruptcy forms -- including a fraudulent affidavit for users to sign claiming that they had done all the legal research on their own.

When a bankruptcy trustee noticed errors in forms submitted by Jayson Reynoso, he blamed the AI website and Ihejirika joined him in a federal court hearing.

The judge ruled that Ihejirika had committed fraudulent, unfair, or deceptive conduct through his computer program, and had effectively engaged in the unauthorised practice of law.

Ihejirika was fined, banned from offering the same service in the future, and ordered to give up the fees he'd collected from nine customers. He appealed but the appeal court upheld the ruling.

This represents a milestone in the history of computing -- the first, admittedly unlicensed, cyberlawyer. As AIs proliferate throughout the legal profession, regulators will have to find some way of dealing with a development that isn't going to just go away.