Your AI Hiring Vendor Is Now a Legal Co-Defendant. Act Accordingly.

The assumption that has quietly governed HR tech procurement for the past decade goes something like this: the vendor builds the tool, the employer deploys it, and legal exposure lives on the employer's side of the table. A California federal court just stress-tested that assumption — and it didn't hold.

A judge refused to dismiss most bias allegations against Workday in a case that keeps FEHA and ADA claims alive against the HR software giant itself, not merely against the employers using its platform. The ruling treats Workday as a potential agent in the hiring process — a legal framing that should make every procurement conversation with an AI hiring vendor feel materially different starting now. If the vendor's algorithm is making or shaping decisions about who gets screened in or out, the vendor may not be an arm's-length tool supplier. It may be a co-decision-maker.

This matters beyond Workday. The entire architecture of AI-powered recruiting — resume parsing, fit scoring, automated screening — rests on models trained on historical hiring data. That data carries the biases of every organization that generated it. AI tools can miss nontraditional candidates whose experience doesn't map cleanly onto credential-matching logic: career-break returners, self-taught technologists, candidates from non-pedigreed institutions. A SHRM26 speaker made the point plainly: employers must actively design AI systems to recognize experience that falls outside conventional patterns, not assume the vendor has already solved for it. Most haven't.

The harder problem is structural. The CHRO's impossible mandate right now is to accelerate AI adoption while simultaneously managing the risk surface that adoption creates. Most organizations haven't built the governance muscle to hold both at once — so the CHRO absorbs the contradiction personally, acting as both accelerant and brake. That's not a tenable position when legal exposure is sharpening.

What the Workday ruling actually demands is a harder look at vendor contracts, audit rights, and what "explainability" commitments actually mean in practice. When a candidate is rejected by an AI screening tool, can your organization reconstruct why? Can your vendor? If the honest answer to either question is no, you're operating a black-box hiring process and calling it efficiency. That's a defensible choice only until it isn't.

The returnship angle makes this more pointed. Employers who invested in returnship programs are now pulling back, according to the Financial Times — a rollback that could deepen unemployment for people returning from career breaks. The irony is brutal: companies are simultaneously abandoning the human programs designed to reintegrate nontraditional candidates and deploying AI systems that, left unchecked, will screen those same candidates out automatically. The gap between stated DEI commitments and actual hiring architecture has never been easier to document — or litigate.

The AI atrophy risk runs in parallel. CIOs at the 2026 MIT Sloan CIO Symposium named the erosion of critical thinking skills as a direct byproduct of over-reliance on AI tools. In recruiting specifically, this means the human judgment that should be the check on algorithmic error — the recruiter who notices that a candidate's unconventional background is actually a signal, not noise — atrophies fastest in the organizations that automate most aggressively. You don't just risk bias. You risk losing the organizational capacity to detect it.

None of this argues against AI in hiring. It argues against AI in hiring without governance infrastructure that matches the legal and human stakes. The Workday case is an early data point, not an endpoint. More litigation is coming, and the plaintiffs' bar is watching carefully to see which theories survive dismissal.

The question every CHRO should bring to the next vendor review meeting is not "what does this tool do?" It is "when this tool causes a disparate impact finding, what is our audit trail, and where does your liability end and ours begin?" If the vendor can't answer that question with contractual specificity, the efficiency gains in the sales deck don't justify the risk exposure on your balance sheet.

AI-powered hiring isn't the problem. Buying AI-powered hiring like it's infrastructure procurement — commodity, low-scrutiny, set-and-forget — is.