California’s New AI Crackdown on the Workplace

Context

California lawmakers are back with a clear message: no “robo bosses,” and no surprise AI layoffs.

Two new Senate bills introduced on February 2 would significantly tighten how employers use artificial intelligence in workforce decisions. Together, they would require human review of AI-driven terminations and extend mass layoff notice periods when automation is involved.

If passed, these would be among the strictest AI employment regulations in the country.

Why It Matters

For business leaders, this is not just a California story.

The state often sets the regulatory tone for the rest of the country. Multistate employers should treat these proposals as an early warning signal. The implications touch compliance, documentation, technology procurement, workforce planning, and risk exposure.

AI in HR is no longer just an efficiency play. It is becoming a regulatory flashpoint.

Core Idea

AI can assist managers. It cannot replace them.

Both bills are built on the same principle: artificial intelligence may inform employment decisions, but it cannot be the final decision-maker, and it cannot quietly eliminate jobs without extended warning.

SB 947: The Return of the “No Robo Bosses” Act

SB 947, introduced by Senator Jerry McNerney, revives a bill Governor Newsom vetoed last year, with several key revisions.

At its core, the bill would prohibit employers from using an “automated decision system” as the sole basis for discharging or disciplining a worker. The definition is broad, covering any computational process, including machine learning or data-driven systems, that makes or facilitates employment decisions.

Key updates from last year:

• Advance notice removed: Employers would no longer need to notify workers before deploying AI systems. This addresses a central concern raised in last year’s veto.
• Retroactive notice required: Employers must notify workers after AI has been used to make or substantially assist in a discharge or discipline decision.
• Predictive AI restricted: New language would prohibit systems that attempt to predict future behavior, such as tools forecasting performance, misconduct, or turnover risk.
• Human decision required: A “natural person” must make the final call in termination or discipline decisions. AI cannot be the sole decision-maker.
• Private enforcement risk: The Labor Commissioner could impose penalties, and workers or their representatives could bring civil actions, including claims for punitive damages.

This is not symbolic oversight. It creates real exposure if human review is superficial or poorly documented.

SB 951: 90-Day Notice for AI-Driven Layoffs

SB 951, introduced by Senator Eloise Gómez Reyes, targets automation-related mass layoffs by amending California’s Cal-WARN Act.

The proposal would:

• Extend notice to 90 days: AI- or automation-driven mass layoffs would require 90 days’ advance notice, 30 days longer than current law.
• Lower the trigger threshold: The requirement would apply at 25 employees or 25 percent of the workforce, whichever is less, compared to the standard 50-employee threshold.
• Create internal job bidding rights: At employers with 100 or more employees, affected workers could bid on open positions during the notice period.
• Mandate state reporting: Employers would need to notify the state about AI-driven layoffs, creating a formal database of automation-related job losses.
• Expand coverage: The bill would apply to employers with 75 or more employees.

For employers implementing automation initiatives, this fundamentally changes the timeline. Workforce reductions tied to AI would need to be planned months in advance.

Key Actions for Employers Now

Even before passage, leaders should prepare.

• Audit your AI systems: Inventory all tools that could qualify as automated decision systems, including applicant tracking systems, performance platforms, workforce analytics, predictive models, and scheduling algorithms. Do not rely on vendor marketing labels. The statutory definition is expansive.
• Build real human oversight: Document meaningful human review in any termination or discipline decision where AI plays a role. Rubber-stamping AI outputs will not withstand scrutiny.
• Prepare disclosure protocols: If retroactive notice becomes law, termination processes and documentation will need to explain how AI was used and how the final decision was made.
• Revisit workforce planning timelines: For any automation initiative that could lead to job losses, model a 90-day notice requirement into your implementation roadmap.
• Assess litigation exposure: Private rights of action and potential punitive damages increase the stakes. AI governance must be integrated with employment law compliance.

The Political Wild Card

The legislative path is uncertain.

Labor groups are prioritizing these measures after losing ground last year. The revisions to SB 947 appear designed to address Governor Newsom’s prior objections.

At the same time, the tech industry is likely to argue that the bills impose heavy compliance burdens and undermine California’s competitiveness. Federal pressure to limit state-level AI regulation could further complicate enforcement.

This debate is not just about HR practices. It is about who sets the rules for AI in the workplace.

Closing Thought

AI is transforming how companies hire, manage, and scale. California is moving to ensure it does not also become the silent decision-maker behind pink slips.

For employers, the takeaway is simple: if AI touches employment decisions, governance cannot be an afterthought. Human judgment, documentation, and strategic planning must lead.