When Joanna Frivet’s husband ran into a legal problem while travelling abroad, he turned to an artificial intelligence (AI) chatbot rather than disturb his wife, who is a lawyer, in the early hours of the morning in Ireland.
Half an hour later, he discovered that the bot had confidently cited a nonexistent law. It was not malicious. It happened because the bot had been trained on a broad data set and was unable to narrow its focus to deal with his specific problem.
Reflecting on her husband’s experience, Frivet began to realise just how poorly general AI performs in terms of legal accuracy.
Some law firms have already banned its use because inaccurate information has led lawyers to cite non-existent case law in court.
Frivet felt s
Continue Reading on The Irish Times
This preview shows approximately 15% of the article. Read the full story on the publisher's website to support quality journalism.