As we previously reported here and here, employers must navigate a rapidly evolving legal landscape as artificial intelligence (AI) continues to transform the modern workplace. From federal rollbacks to aggressive state-level regulation, the use of AI in employment decisions—particularly in hiring, performance management, and surveillance—has become a focal point for lawmakers, regulators, and litigators alike. This article contains an overview of the shifting federal landscape on AI at work, the state level response, and offers recommendations for employers to mitigate risk.

Trump Administration Rolls Back Federal AI Oversight

The federal approach to AI in employment has undergone a dramatic shift in 2025. On his first day in office, President Trump rescinded Executive Order 14110, which directed federal agencies to address AI-related risks such as bias, privacy violations, and safety concerns. This was followed by the removal of key guidance documents from the U.S. Equal Employment Opportunity Commission (EEOC), including technical assistance on Title VII compliance and the Americans with Disabilities Act (ADA) as they relate to AI tools. The Department of Labor has also signaled that its prior guidance on AI best practices may no longer reflect current policy, leaving employers with less clarity at the federal level than ever.

Despite these reversals, employers remain liable under existing anti-discrimination laws for the outcomes of AI-driven employment decisions—even when those tools are developed by third-party vendors.

States, Especially California, Fill the Federal Void With AI Regulation

In the absence of clear federal guidance, states have begun regulating AI in the workplace. California, in particular, has emerged as a bellwether.

Earlier this year, California introduced several bills aimed at curbing the unchecked use of AI in employment decisions:

• SB 7 – “No Robo Bosses Act”: This bill would require employers to provide 30 days’ notice before using any automated decision system (ADS) and mandates human oversight in employment decisions. It also bans AI tools that infer protected characteristics or retaliate against workers.
• AB 1018 – Automated Decisions Safety Act: This legislation would impose broad compliance obligations on both employers and AI vendors, including bias audits, data retention policies, and impact assessments.
• AB 1221 and AB 1331: These bills target AI-driven workplace surveillance, requiring transparency and limiting monitoring during off-duty hours or in private space.

Other states are following suit. For example, Illinois, Colorado, and New York City already have laws regulating AI in hiring, and over 25 states have introduced similar legislation in 2025.

Practical Implications for Employers

With limited federal guidance and state laws multiplying, inaction is risky and, indeed, reckless. Employers should therefore take proactive steps to mitigate legal exposure:

1. Audit AI Tools Regularly: Conduct bias audits and impact assessments to ensure compliance with anti-discrimination laws.
2. Review Vendor Agreements: Ensure contracts with AI vendors include provisions for transparency, data handling, and liability.
3. Train HR and Leadership: Equip decision-makers with the knowledge to use AI responsibly and in compliance with applicable laws.
4. Implement Human Oversight: Avoid fully automated employment decisions. Ensure a human reviews and approves critical outcomes.
5. Stay Informed: Monitor legislative developments in all jurisdictions where your business operates.

Looking Ahead: Balancing Innovation With Worker Protection

The use of AI in the workplace is not going away. In fact, it is accelerating. But with innovation comes responsibility. Employers should systematically document data collection from workplace technologies, update privacy-related polices, and ensure human oversight is integrated into any employment decisions that rely on algorithmic input.