AI tools have found their way into HR workflows at companies of every size—screening resumes, scoring video interviews, flagging performance issues, and recommending terminations. They promise speed and objectivity. What they often deliver is undetected bias, compliance exposure, and legal liability that most employers never see coming. If your HR department is using AI to make or influence employment decisions, a New York employment lawyer can help you assess your risk before a complaint lands on your desk.
AI in Hiring Is Already Regulated in New York City
New York City enacted Local Law 144, which requires employers and employment agencies using automated employment decision tools (AEDTs) in hiring or promotion decisions to complete an independent bias audit before deploying those tools—and to repeat that audit annually. The law also requires employers to notify candidates and employees that AI is being used in the evaluation process.
The New York City Department of Consumer and Worker Protection (DCWP) enforces the law, and a December 2025 audit by the New York State Comptroller’s Office found that enforcement has faced challenges, with criticism that may intensify future scrutiny of employers who are not in compliance.
One critical point: human review does not necessarily remove the tool from the law if the AEDT substantially assists the decision. If an AI tool ranks, scores, or filters candidates and a human uses that output to reach a decision, the employer is still subject to Local Law 144. Human oversight does not eliminate the compliance obligation—it just adds a layer of documentation you need to have in place.
The Federal Anti-Discrimination Exposure Is Real
Beyond city and state requirements, employers using AI in HR decisions face exposure under federal anti-discrimination law. Title VII of the Civil Rights Act, the Americans with Disabilities Act, and the Age Discrimination in Employment Act all apply when AI tools influence hiring, discipline, or termination.
The problem is structural: AI systems are trained on historical data, and if that data reflects past discriminatory patterns, the tool can reproduce those patterns at scale without anyone flagging it as intentional discrimination.
Courts and regulators do not require proof of intent. If an AI tool produces a disparate impact on a protected class—screening out a disproportionate number of applicants over 40, for example, or rejecting candidates with certain disability-related employment gaps—that outcome can support a discrimination claim regardless of whether the employer meant to discriminate.
In 2024, AI-powered hiring tools processed over 30 million applications and triggered hundreds of discrimination complaints. That number is growing.
What Employers Get Wrong About Vendor Liability
A common assumption is that if the AI tool comes from a vendor, the vendor bears responsibility for what it does. That assumption is wrong. When an employer deploys an AI tool to make or assist in employment decisions, the employer remains legally responsible for those decisions. Vendor contracts frequently disclaim liability for discriminatory outcomes, leaving employers to defend claims that originate from software they did not build and may not fully understand.
Before adopting any AI-driven HR tool, employers should ask vendors directly: Has this tool been audited for bias? What protected classes were tested? What does the audit show about disparate impact? How is the tool updated, and what triggers a new audit? If a vendor cannot answer those questions clearly, that is a signal to pause before deployment.
The Regulatory Landscape Is Accelerating
New York City’s law is not an outlier—it is the front edge of a wave. Illinois now requires employers to notify workers when AI is used in any employment decision, effective January 1, 2026, and prohibits discriminatory outcomes regardless of intent. New Jersey has reinforced its Law Against Discrimination to cover automated decision tools, clarifying that disparate impact claims apply to AI-driven outcomes. California’s Automated Decision-Making Technology rules took effect on October 1, 2025, establishing some of the most detailed anti-discrimination standards in the country for AI systems used in employment.
Federal legislation may follow, but for now employers face a patchwork of state, local, and federal anti-discrimination obligations. Employers who have not yet examined how AI fits into their hiring and HR practices are falling behind a compliance curve that is only getting steeper.
How Orin Kurtz Helps Employers Manage AI Risk
Attorney Orin Kurtz has spent more than 20 years advising New York employers on the legal dimensions of workplace decisions—including those HR teams make every day. As AI becomes embedded in those processes, the legal stakes rise.
Orin works with employers to review how AI tools are being used, assess exposure under city, state, and federal law, and develop compliance practices that reduce the risk of audits, agency investigations, and litigation. If a complaint has already been filed, he provides continuity from preventive counsel to defense.
Acting before a problem surfaces is almost always less expensive than responding after one does. Contact Orin Kurtz to discuss how AI use in your HR department may be affecting your legal exposure.
Frequently Asked Questions
Does NYC Local Law 144 apply to small businesses?
Yes. The law applies to employers and employment agencies using automated employment decision tools in hiring or promotion decisions in New York City, regardless of company size. There is no small-business exemption.
Can we use AI in HR if a human reviews all final decisions?
Not without compliance obligations. NYC’s law applies whenever an AI tool influences the decision, even if a human signs off at the end. The DCWP has confirmed that human review does not eliminate the bias audit or notification requirements.
What should we do if our current HR software uses AI?
Start by asking your vendor for documentation on bias audits and disparate impact testing. Then consult with an employment attorney to assess whether your current use triggers any city, state, or federal requirements. Getting ahead of this issue now is substantially less costly than responding to a complaint or audit later.