When there have been conversations about how the legal industry can leverage generative artificial intelligence in recent weeks, questions around large language models' safety and data privacy have often followed.