Pankit Doshi discusses California's new AI employment regulations in the Daily Journal

The Daily Journal recently featured Pankit Doshi and Nicholas Neathamer’s insights on how California’s new FEHA regulations governing AI and automated decision systems are reshaping employment litigation, as employers navigate expanded discovery risk under rules that took effect October 1, 2025. Pankit and Nicholas explain that a bias audit is no longer a simple compliance checkbox — courts will scrutinize the quality and follow-through of those efforts, meaning a narrow or shelved audit can do more harm than good — while relying on an audit as a defense can itself trigger a waiver of the privilege meant to protect it, since privilege cannot be used as both a sword and a shield.

The authors point to the closely-watched Mobley v. Workday litigation as an early signal of where this is headed, noting that outsourcing an AI hiring tool doesn’t outsource liability — employers can remain on the hook for a vendor’s discriminatory outputs even when they had no hand in building the underlying model. They also warn that standardized algorithmic tools may erode a traditional defense against class certification, since a uniformly applied algorithm can itself serve as the “common policy” plaintiffs need to satisfy commonality requirements.

The takeaway, as Pankit and Nicholas put it: California’s regulations don’t ban AI in the workplace, but they do turn AI governance into either a litigation asset or a liability — and the question isn’t whether AI-related employment litigation is coming, but that it’s already here.