AI and the Future of Law GPT-5 and the Future of Legal AI Regulation
Sep 2, 2025
Explore whether GPT-5 is ready for legal practice and the vital role of human oversight in the courtroom. The discussion highlights AI's potential to reduce human error and the challenges faced by legal professionals in integrating new technology. Tune in for innovative features of GPT-5, and insights from a pivotal report on consumer-focused AI regulation in legal services. The conversation also touches on the evolving roles of legal professionals and how communities adapt to these advancements.
AI Snips
Chapters
Transcript
Episode notes
Using GPT-5 To Measure Court Error
- Bridget used GPT-5 to quantify tolerated human error in courts and found surprising appellate reversal rates for her jurisdictions.
- She iterated prompts to layer reversals, harmless error, exonerations, and unappealed cases to build persuasive slides and talking points.
Write First, Then Use AI To Refine
- Do the initial thinking and draft with your human brain before using AI to refine or edit.
- Use AI later to improve structure, cross-references, and polish after you’ve written first.
AI Is Middle-To-Middle Support
- Bridget observes AI is 'middle to middle' support, keeping humans essential at the front (framing) and back (QC).
- That framing explains why humans remain valuable despite stronger models.