Decoder with Nilay Patel The surprising case for AI judges
105 snips
Feb 12, 2026 
Bridget McCormack, former Michigan chief justice and now CEO of the American Arbitration Association, leads development of the AI Arbitrator. The conversation covers how an AI-assisted arbitration platform is built, when automation fits disputes, and the tradeoffs around trust, transparency, and risks like hallucinations. It also explains why they started with documents-driven construction cases.
AI Snips
Chapters
Transcript
Episode notes
Determinism Would Reduce Disputes
- The legal system should be more deterministic to reduce disputes and improve planning.
- Bridget McCormack argues AI can make many routine legal outcomes predictable and efficient.
AI Makes Parties Feel Heard
- AI systems can enhance perceived fairness by ensuring parties feel heard and understood.
- McCormack says iterative agent summaries let parties confirm claims and evidence until satisfied.
Start Narrow And Keep Humans In Loop
- Start with narrow, document-only dispute types where evidence is structured.
- Use human arbitrators in the loop to review AI reasoning and finalize awards.