Industries

AI now decides who gets hired. AVAAS certifies the AI that decides.

Hiring is the most heavily automated high-stakes decision in the economy. Resume screeners, ranking models, and assessment tools now stand between applicants and a livelihood. When that AI gets it wrong, the harm lands on a person and the liability lands on the employer and the vendor. AVAAS certifies the system at the moment it screens a candidate.

Title VIIADA / ADEANYC Local Law 144Illinois AIVICalifornia ADMTColorado AI Act
Assess your hiring AI →
Where AI acts on a person

The decision point in hiring

In hiring, the AI acts when it screens a resume, ranks a candidate, scores a video interview, or routes someone out of the pipeline. The moment it does, the law treats that as an employment decision.

When a qualified applicant is filtered out by a model, you owe a defensible answer for why, and a bias audit alone does not provide it.

What keeps you exposed

What keeps employers exposed

Disparate impact

A model screens out a protected group and you cannot see it

Adverse impact emerges from features that look neutral. If you cannot measure it across protected classes, you cannot defend it.

Vendor liability

Your hiring tool was built by someone else

Courts are testing whether an AI vendor can be liable as your agent for discriminatory screening. A vendor assurance is not your evidence.

The audit gap

You passed a bias audit and still got sued

A Local Law 144 audit checks a narrow slice at a point in time. It is the floor, not proof the system behaves in production.

This is already happening
In Mobley v. Workday, a federal court is allowing claims to proceed on the theory that an AI hiring vendor can be liable as an agent for discrimination, and the case is in discovery.
Mobley v. Workday, N.D. Cal. · Kistler v. Eightfold AI, FCRA, pending
How AVAAS adds value

Evidence that the system hires lawfully

Does the model produce adverse impact across protected classes?

Five structurally independent validators test the system for demographic disparity using causal attribution, so impact traces to the features that actually drove it.

Can you explain why a specific candidate was screened out?

AVAAS evaluates whether the system can give a specific, accurate reason for each decision rather than a post-hoc guess.

Is the certified system the one you are running?

Sealed deployment verification confirms the model in production is the one that was certified.

You get documented, third-party evidence that your hiring system does what the law requires, before a regulator or a rejected applicant asks. Because AVAAS is operated separately from the standard it certifies, the validation is genuinely independent.

Related AVAAS coverage: NYC Local Law 144 · California Employment · Illinois · Colorado · California ADMT. Or run the free Regulation Checker to see what applies to you.

See where your hiring AI creates liability.

Tell us how AI touches your hiring, screening, or promotion decisions, and we will scope an AVAAS certification to the exposure.

Ready to start now? Certify Your AI →  or  email [email protected]