Episode Key Takeaways
The employer stays on the hook for every hiring decision, regardless of which tools are deployed. Even if tech vendors face liability under U.S. law, that never fully exonerates the hiring organization. Responsibility for employment decisions is non-transferable.
Risk scales logarithmically as AI moves closer to the actual hiring decision. Using AI to draft rejection emails or write job descriptions is defensible; automating pre-interview sorting of candidates creates exponentially higher legal exposure because you’re automatically excluding qualified people based on opaque pattern-matching.
Heather Bussing emphasizes that the legal frameworks don’t ban AI in recruiting—they ban autonomous decision-making. You can use AI as a tool throughout the process, but a human must review and approve before any employment decision is made. The closer the tool sits to the final choice, the more scrutiny it faces.
Monitoring outcomes and running bias audits is the closest thing to a legal safe harbor. The EU Pay Transparency Act and emerging U.S. frameworks reward organizations that actively search for bias, investigate root causes, involve employees in the fix, and document corrective action—not those that claim perfection.
Everything is now on the record. The shift from plausible deniability to full transparency means hiring decisions must be defensible in writing. Candidates use AI transcription tools, Zoom records everything, and vendors log all interactions—so the old approach of ‘we just liked them’ no longer works.
Frequently
Asked
Questions
Does the EU AI Act apply to U.S. companies?
Yes, if you post a job globally or accept remote applications, you may be subject to EU regulations regardless of where your company is based. The EU AI Act classifies hiring as high-risk and requires human oversight before employment decisions. Other jurisdictions—Colorado, New York City—have similar frameworks. You must know who may apply to your job posting.
Can employers use AI in hiring, or is it banned?
AI is not banned. You can use it as a tool throughout recruiting—writing emails, drafting job descriptions, analyzing resumes. What’s prohibited is letting AI make employment decisions alone. A human must review and approve before any hiring decision is finalized. The closer the tool sits to the final decision, the higher the legal risk.
What should TA leaders present to legal and compliance teams?
Explain how the tool works, what data it relies on, whether the vendor has bias-tested it, and what monitoring you’ll do. Legal teams want to know how you’ll reduce risk, not eliminate it. Show that you’ve vetted the vendor, plan to audit outcomes, and have people in the loop. Avoid pitches framed as ‘we’ll save money by removing people’—you need more oversight, not less.
What happens if an AI tool makes a biased hiring decision?
The employer is liable. However, organizations that proactively audit for bias, investigate problems, involve employees in solutions, and document corrective action have stronger legal defenses. The goal is to show good-faith effort and accountability, not perfection. Causation in AI cases will be complex, but demonstrating a responsible process matters significantly in litigation.
Where should TA professionals find reliable AI hiring guidance?
Zwillgen, a technology law firm, publishes clear, regularly updated guidance on AI and hiring with sources for deeper research. They specialize in this space and are a trusted first stop. Also consult your own employment counsel and vendor documentation. Avoid generic AI advice; seek sources that understand both employment law and technology.