In a seven-exercise IRAC study, general-purpose LLMs outperformed Lexis+ AI, and all tested LLMs showed limitations including brevity, hedging, false confidence, and hallucinations.
The False Promise of ChatGPT
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CY 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Artificial Intelligence and Legal Analysis: Implications for Legal Education and the Profession
In a seven-exercise IRAC study, general-purpose LLMs outperformed Lexis+ AI, and all tested LLMs showed limitations including brevity, hedging, false confidence, and hallucinations.