“Disturbing & pervasive errors among three popular models on a wide range of legal tasks” study found
“hallucination rates range from 69% to 88% in response to specific legal queries for state-of-the-art language models”
“performance deteriorates when dealing with more complex tasks that require a nuanced understanding of legal issues or interpretation of legal texts”
“case law from lower courts … subject to more frequent hallucinations than case law from higher courts
“model susceptibility to what we call “contra-factual bias,” namely the tendency to assume that a factual premise in a query is true, even if it is flatly wrong”
https://hai.stanford.edu/news/hallucinating-law-legal-mistakes-large-language-models-are-pervasive