The Pat Kenny Show What happens when AI is used in court?
Dec 1, 2025
Tomás Keyes, a barrister and legal commentator, explores the intriguing intersection of AI and law. He discusses the promise and risks of using AI in legal submissions, emphasizing the importance of informed use. Keyes sheds light on AI hallucinations, where generative tools create plausible but false information. He recounts the MATA case, where ChatGPT's fabricated citations led to a judge's inquiry. Keyes also covers the necessary guidance and potential sanctions for misuse of AI, offering vital insights for legal professionals navigating this new frontier.
AI Snips
Chapters
Transcript
Episode notes
AI Hallucinations Can Mislead Courts
- AI can produce plausible-sounding but false legal information, a phenomenon called 'hallucination'.
- Tomás Keyes warns that new AI power requires careful understanding and verification by lawyers.
Always Verify AI Outputs With Trusted Sources
- Verify AI-generated legal content against authoritative legal databases and textbooks.
- Use professional legal research tools rather than relying solely on general AI outputs.
LLMs Prefer Plausible Answers Over Admitting Ignorance
- Tomás Keyes likens AI hallucinations to an unprepared student eager to please.
- Large language models produce plausible answers rather than admitting uncertainty.