A U.S. judge appeared to have misquoted case law using artificial intelligence (AI) in handling a lawsuit involving Elon Musk, the Tesla chief executive officer (CEO). The judge had previously issued a ruling favorable to Musk.
According to the San Francisco Chronicle on the 20th (local time), Aaron Greenspan, who filed a lawsuit accusing Musk of defamation and securities fraud, claimed that Judge Joseph Quinn's recent order contained what appeared to be an AI error.
Greenspan, an investor who shorted Tesla stock, filed suit after criticizing Tesla's "Full Self-Driving" (FSD) as exaggerated and Musk labeled him a conspiracy theorist.
In response, Musk filed a motion to strike under a California state law mechanism that allows for early dismissal of lawsuits aimed at suppressing free speech.
Greenspan argued that Musk filed the motion after the 60-day court deadline, but Judge Quinn cited appellate case law allowing the filing date to be applied retroactively and dismissed most of Greenspan's claims. He also ordered Greenspan to pay the litigation expenses.
The problem is that the precedent Quinn cited was wrong. The part about allowing retroactive filing was not a judicial determination but an argument by the defense, and it was rejected in the final ruling.
On this point, the legal outlet "Above the Law" noted that a typical AI error known as "hallucination" occurred, and it appeared the judge cited it without review.
In the end, Judge Quinn rewrote the order by hand to correct it, but the order itself reportedly remained in place.
According to the San Francisco court, judges are allowed to use AI, but case law must ultimately be reviewed by a human.