Tech Tips

• Citation: Vol. 47 No. 1 Pg. 52
• Publication year: 2024
• Pages: 52

Tech Tips Vol. 47 No. 1 Pg. 52 Wyoming Bar Journal February 2024

Blake A. Klinkner

Blake Klinkner is a professor at the University of North Dakota School of Law in Grand Forks, North Dakota.

Why Are Artificial Intelligence Chatbots (Presently) So Bad at Legal Research and Writing

Artificial intelligence became much more popular—and accessible—in 2023 as various artificial intelligence platforms were released for widespread public consumption, most notably OpenAI’s ChatGPT chatbot. Artificial intelligence chatbots may now be accessed by the public easily, and for free, and chatbot users have been generally amazed at the kinds of responses these chatbots may presently provide to user queries. However, 2023 showed the legal community that artificial intelligence chatbots are not presently capable of engaging in reliable legal research and drafting, as evidenced by some headline-making cases throughout the country where attorneys were sanctioned for filing legal briefs which contained contained made-up cases and phony legal doctrines generated by artificial intelligence chatbots.^[1] Why are artificial intelligence chatbots presently so bad at legal research and writing?

What are Artificial Intelligence “Hallucinations?”

Currently, the main problem with relying upon artificial intelligence to perform legal research and writing is that artificial intelligence is prone to “hallucinating.” Artificial intelligence “hallucinations hallucinations,” simply stated, are when artificial intelligence...