In Redo of Its Study, Stanford Finds Westlaw’s AI Hallucinates At Double the Rate of LexisNexis

The study conducted by Stanford University researchers found that generative AI legal research tools from LexisNexis and Thomson Reuters produce incorrect results more frequently than claimed in their marketing efforts.

Thomson Reuters initially refused access to their AI-Assisted Research product but later agreed, and the updated study found it to perform unfavourably compared to the LexisNexis product.Westlaw’s AI-assisted research yielded accurate responses only 42% of the time, compared to 65% for Lexis+ AI.

Westlaw was found to produce incorrect results at almost twice the rate of the LexisNexis product.The study emphasised the importance of rigorous and transparent benchmarking and public evaluations of AI tools in the legal field.

Source: Law Sites