Using AI in Hiring? Answer These Compliance Questions First

Resume screeners and assessment tools are the employer's responsibility even when a vendor built them. Here are the compliance questions to answer before you rely on AI in hiring.

AI tools that screen resumes, rank candidates, and score assessments promise to save hiring teams a great deal of time. They also introduce a category of compliance risk that many employers underestimate. When an automated tool influences who gets hired, the legal responsibility for the outcome rests with the employer, not the software vendor.

That single principle should shape how you evaluate any AI hiring tool. Here are the questions to answer before you rely on one.

The vendor built it, but you are accountable

It is tempting to assume that a purchased tool has already handled the compliance questions. It has not, at least not on your behalf. If a screening tool produces results that disadvantage a protected group, the employer using it can be held responsible. Outsourcing the software does not outsource the obligation.

Treat any AI hiring tool as an extension of your own hiring process, subject to the same standards you would apply to a human decision. The same is true of the notices and disclosures candidates are entitled to. If a tool influences who advances, applicants may have a right to know it is being used, and that responsibility sits with you as the employer rather than with the vendor who supplied the software.

The regulatory landscape is moving

The rules around automated hiring are developing, and a few reference points are already firmly in place.

• New York City's Local Law 144 on automated employment decision tools took effect with enforcement beginning in July 2023, requiring bias audits and notice to candidates.
• The EEOC has stated that existing federal anti-discrimination law applies to employers' use of AI tools.
• Several states are moving to regulate the use of AI in employment decisions, so the requirements you face may expand over time.

Even where a specific local law does not yet apply to you, the direction is clear, and existing anti-discrimination law already does.

Ask vendors for bias-audit documentation

Before adopting a tool, ask the vendor directly for documentation of how it has been tested for bias. A credible vendor will be able to explain what was audited, how, and by whom. Vague reassurance is not documentation. If a vendor cannot produce meaningful evidence, treat that as a significant warning sign.

Keep copies of what the vendor provides. If your process is ever questioned, that documentation is part of how you demonstrate you acted responsibly.

Keep a human decision in the loop

An AI tool should inform hiring decisions, not make them. Keeping a qualified person accountable for the final call is both good practice and an important safeguard. It gives you a point where context, judgment, and fairness can be applied, and it prevents an opaque automated ranking from quietly becoming the decision.

Make clear internally who owns the decision and that the tool is one input among several.

Document your process

Whatever tools you use, keep a record of how your hiring process works, what the tool does, what notice candidates receive, and how you verified the tool. Documentation is what turns a defensible intention into a defensible record. It is also what regulators increasingly expect employers to be able to produce.

Practical next steps

Confirm which laws apply where you hire, request bias-audit documentation from any vendor, ensure a qualified person makes the final decision, and document the whole process including candidate notice. A managed services partner can help you sequence this work so your hiring stays both efficient and compliant.