AI in Hiring – Proceed with Caution

The rapid adoption of artificial intelligence in employment decisions presents significant opportunities for efficiency—but also substantial legal and operational risks. Whether your organization uses AI to screen resumes, rank candidates, or conduct preliminary interviews, understanding the evolving regulatory landscape and implementing robust safeguards is essential to protect your business.

Employers must recognize that long-standing federal and state antidiscrimination statutes—including Title VII of the Civil Rights Act, the Americans with Disabilities Act (ADA), and the Age Discrimination in Employment Act (ADEA) and the Texas Labor Code—apply fully to outcomes produced by AI systems. Courts evaluate AI-driven hiring decisions using established disparate impact frameworks, focusing on statistical disparities, business necessity, and the availability of less discriminatory alternatives. Recently, a California federal court allowed disparate impact claims to proceed against an AI vendor, reasoning that "nothing in the language of the federal anti-discrimination statutes . . . distinguishes between delegating functions to an automated agent versus a live human one."

State legislatures are increasingly enacting AI-specific statutes. The Texas Responsible AI Governance Act (TRAIGA), effective January 1, 2025, prohibits AI systems from being developed or deployed to unlawfully discriminate against a protected class. Similar legislation has emerged in Illinois, Colorado, and California, imposing notice requirements, bias audits, and documentation obligations on employers using automated decision-making tools. Employers operating in multiple jurisdictions face a complex patchwork of compliance requirements.

AI systems can generate discriminatory outcomes through several mechanisms: biased training data reflecting historical inequities, model design choices that inadvertently penalize certain groups, and other variables (such as zip code or graduation date) that may correlate with protected characteristics. Because many AI models are opaque— producing decisions through layers of numerical weights rather than transparent rules— employers may struggle to identify bias before harm occurs. Research confirms these concerns: a University of Washington study found that AI models preferred resumes with white-associated names 85% of the time, and when human reviewers received biased AI recommendations, they mirrored that bias in their own selections.

Risk Mitigation Strategy

To minimize legal and operational exposure, employers should adopt a structured governance framework:

• Maintain Human Oversight. AI should inform—not replace—human judgment. Ensure that qualified personnel review AI recommendations before any employment decision becomes final, and document the rationale for each hiring outcome.

• Secure Favorable Vendor Agreements. When using third-party AI tools, negotiate indemnification provisions that allocate risk appropriately. Employers remain responsible for discriminatory outcomes even when the underlying system is developed by an external provider.

• Establish Ongoing Monitoring. AI systems evolve through model updates and retraining. Implement periodic audits and governance protocols to ensure continued compliance as both the technology and regulatory requirements change.

The use of AI in hiring is not inherently unlawful, but it requires the same level of legal oversight as traditional selection procedures. Employers that treat AI-assisted decision-making as a core employment practice—subject to rigorous auditing, documentation, and human review—will be best positioned to capture the benefits of automation while minimizing liability.