AI Can’t Handle the Truth When it Comes to the Law

HAI Senior Fellow Daniel E. Ho shows large language models hallucinate frequently when used for legal queries.