Our Take
This is a compliance clock, not a crisis: the order imposes zero immediate obligations but forces California's labor agency to recommend changes to layoff notice rules within six months, giving employers a narrow window to audit their current practices before the hammer falls.
Why it matters
California HR leaders operate under real risk. Connecticut and New York have already tightened WARN laws to flag AI-driven layoffs separately; California's recommendations could be stricter, and SB 951 pending in the state legislature already proposes 90-day notice (vs. 60 today) for automation-driven cuts and a lower 25-worker threshold. The sooner you know what your AI is actually doing in hiring, firing, and performance reviews, the sooner you can build defensible practice.
Do this week
HR leaders: conduct an audit of every AI tool or algorithm in your employment lifecycle (hiring, performance reviews, severance decisions, hiring freezes) before October 15 so you can flag compliance gaps before California's LWDA recommendations reshape the playing field.
Newsom sets 180-day deadline for AI layoff notice overhaul
California Gov. Gavin Newsom signed an executive order last month directing the state's Labor and Workforce Development Agency (LWDA) to recommend changes to the state's WARN Act by mid-April 2025. The order itself imposes no new compliance obligations immediately, but it sets a hard deadline on what could become binding law within months of those recommendations landing.
The LWDA must study how to revise WARN Act notice requirements for AI- and automation-driven layoffs, and also review severance practices, equity compensation, and displaced-worker support programs, including how other countries handle AI job loss. In parallel, Connecticut and New York have already amended their WARN laws to require separate notice when layoffs stem from AI adoption, giving California a policy template.
The order also directs California's Employment Development Department to launch a public dashboard within 90 days tracking AI's effects on employment by sector, with mandatory employer feedback reports twice yearly through 2027. A second mandate requires study of how union contracts are handling AI adoption by October 15, raising the likelihood that collective bargaining obligations around automation will face scrutiny.
SB 951 is already waiting in the legislature
Newsom's order does not exist in a vacuum. Senate Bill 951, under separate consideration in the California legislature, would amend the Cal-WARN Act to require 90 days' notice for AI- or automation-driven layoffs (30 days longer than standard WARN) and lower the trigger threshold from 50 employees to just 25 workers or 25% of the workforce. The bill would also require employers to flag hiring freezes tied to AI and grant affected workers the right to bid on other open positions.
The executive order's 180-day clock for LWDA recommendations means that statutory proposals could follow quickly. If the LWDA's recommendations align with or exceed SB 951's terms, employers face a compressed timeline to understand what AI systems are actually doing in their hiring, performance management, and reduction-in-force decisions. The order explicitly reaffirms existing California protections against discrimination from automated decision-making tools in hiring, performance reviews, and terminations, signaling that compliance scrutiny will intensify.
Audit AI hiring and employment tools now
Employers using AI anywhere in the employment lifecycle should treat the next six months as a compliance-hardening window, not a waiting period. Document which AI tools or algorithms influence hiring decisions, performance evaluations, severance decisions, or workforce reduction recommendations. Review whether your current WARN Act notification procedures account for AI-driven decisions separately, and whether your severance or equity compensation practices are defensible under potential new state standards.
The order also directs the Governor's Office of Business and Economic Development to study barriers to employee ownership structures and ways workers might share in AI-driven productivity gains through equity. This is early-stage thinking but worth monitoring as a possible legislative model for offsetting displacement concerns. For now, the actionable priority is straightforward: map your AI use, measure your WARN Act readiness, and build a defensible narrative about how and why AI influences employment decisions before California's LWDA recommendations reshape the compliance floor.