Judge blasts lawyer after AI-generated errors derail Australian murder case

A senior lawyer in Australia has apologized to the Supreme Court of Victoria after filing court documents in a murder case that contained fake quotes, made-up laws, and nonexistent case citations, all generated by artificial intelligence.

Defense lawyer Rishi Nathwani KC admitted “full responsibility” for the AI blunder, which delayed a ruling in the case of a 16-year-old accused of murder. The teen, who cannot be identified for legal reasons, was ultimately found not guilty by reason of mental impairment over the April 2023 killing of a 41-year-old woman in Abbotsford.

“We are deeply sorry and embarrassed for what occurred,” Nathwani told Justice James Elliott. The judge, who had planned to conclude the matter that day, was forced to postpone his decision by 24 hours after discovering the faulty submissions.

“At the risk of understatement, the manner in which these events have unfolded is unsatisfactory,” Elliott said on August 14, revealing that the revised submissions still contained fictional legislation created by AI. “Use of AI without careful oversight of counsel would seriously undermine this court’s ability to deliver justice.”

AI errors cause chaos in courtroom murder case

According to court documents, Nathwani and junior barrister Amelia Beech had not properly reviewed their filings before submitting them.

The AI-generated mistakes included: Citing case judgments that don’t exist, misquoting a parliamentary speech, and referencing laws that were never passed.

Prosecutors, who used the defense submissions as a basis for their own, also failed to verify the information, leading to both sides presenting faulty legal arguments.

The boy, who prosecutors alleged had conspired to kill the woman to steal her car and start an “anti-communist army,” was found to have been suffering from untreated schizophrenia and grandiose delusions at the time of the offense. He will remain under supervision in a youth justice facility.

The case adds to a growing list of AI-related legal blunders worldwide, as courts grapple with technology’s potential, and its risks, in the justice system.