The American Arbitration Association recently announced it’s launching an AI-powered arbitrator in November. Many litigators think this future will never arrive in litigation. That litigation requires empathy, gut instinct, and the ability to read a room, things AI can’t do (at least yet).

But what if the decision-maker itself becomes an AI bot?

I spoke with insurance GCs about what cases might be ripe for AI decision-making. They believe an AI decions maker in low-stakes disputes where litigation costs exceed the exposure might work. One GC told me she’d happily refer any case under $50k to an AI arbitrator “in a heartbeat.”

But there’s several issues: Bad faith liability. Algorithmic bias. Questions about appeals and transparency. And the biggest concern: consent.

Here is my post for Above the Law.