The rejection email arrives before you’ve even closed the browser tab. No recruiter name, no feedback. Just a form message saying the company is “moving forward with other candidates.” If that sounds familiar, you may have been screened out by an automated hiring tool before any human ever read your name. What felt like a hiring decision was an algorithm scoring you out of contention in seconds.
Two developments from late 2025 changed what workers in New York can do about it. A December 2025 state audit exposed serious gaps in how the city enforces its own AI hiring law, and Governor Hochul signed legislation that same month strengthening anti-discrimination protections statewide. Understanding both matters for any New York worker who suspects an algorithm cost them a job opportunity. At Schwartz Perry & Heller LLP, we’ve spent more than three decades handling employment discrimination claims in New York, and we’ve been tracking how these developments reshape the legal landscape for workers.
How AI Now Decides Who Gets Hired in New York
Automated Employment Decision Tools (commonly called AEDTs) are computational processes derived from machine learning, statistical modeling, data analytics, or artificial intelligence that issue a score, classification, or recommendation that substantially assists or replaces human hiring judgment. That definition, drawn directly from New York City’s Local Law 144, covers a wide range of tools already in common use: resume-screening software that filters applicants before a recruiter looks at the pile, video interview platforms that analyze facial expressions and vocal tone to generate a candidate score, and predictive ranking systems that sort hundreds of applicants into a short list based on patterns in historical data.
The bias risk in these systems isn’t hypothetical. When an AEDT is trained on a company’s historical hiring data, it learns to replicate whatever patterns existed in that data, including discriminatory ones. Amazon’s abandoned resume-screening tool is the most cited example: the system downgraded resumes containing words associated with women because the historical hiring data it learned from reflected a male-dominated workforce. The tool didn’t need discriminatory intent to produce a discriminatory result. That’s the core problem with algorithmic hiring, and it’s exactly what New York’s legal framework is now built to address.
What NYC’s Bias-Audit Law Actually Requires
Local Law 144, which took effect in July 2023, applies to any employer or employment agency using an AEDT to evaluate candidates for jobs located in New York City. Before deploying such a tool, the employer must commission an independent bias audit, a review conducted by a third party with no financial stake in the tool’s outcome. That audit must calculate selection rates and impact ratios broken down by race, ethnicity, and sex, and it can’t be more than one year old when the tool is used.
The law also requires transparency before evaluation begins:
- Published audit summaries: Employers must post the bias audit results, including the data source and distribution date, on their public-facing website before the AEDT evaluates any candidate.
- Advance written notice: Candidates must receive written notice at least ten business days before an AEDT will be used to evaluate them.
- Data disclosure: Employers must make available information about the type of data the tool collects, its source, and their data retention policy, either on their website or within 30 days of a written request from a candidate.
On paper, these requirements give New York City workers meaningful transparency. In practice, the December 2025 audit findings tell a more complicated story.
The 2025 Audit Revealed Enforcement That Wasn’t Working
The New York State Comptroller’s Office released an audit on December 2, 2025, reviewing how the Department of Consumer and Worker Protection (DCWP) had enforced Local Law 144 from July 2023 through June 2025. The findings were troubling.
Seventy-five percent of test calls placed to the NYC 311 hotline about AEDT issues were never properly routed to DCWP. Workers trying to report a potential violation through the city’s primary public complaint channel were, in most cases, hitting a dead end they had no way of knowing about. The compliance review results were equally stark: examining the same 32 companies’ published bias audits that DCWP had already reviewed, Comptroller auditors identified at least 17 potential compliance issues. DCWP had flagged one.
This matters for how much weight a published bias audit should carry. When an employer points to a posted audit as evidence of compliance, the Comptroller’s findings suggest the agency responsible for verifying that compliance has been catching only a fraction of actual problems. A published audit is a starting point for scrutiny, not a clean bill of health.
New York’s Disparate Impact Protections Just Got Stronger
The second major development from late 2025 applies across all of New York State, not just New York City. On December 19, 2025, Governor Hochul signed S8338, amending the New York State Human Rights Law to expressly recognize disparate impact claims. For workers dealing with AI hiring discrimination, that’s significant.
Disparate impact theory allows a discrimination claim without proving discriminatory intent. A worker needs to show that a neutral-seeming practice (such as an AI screening tool) actually produced or predictably produces a discriminatory effect on people in a protected class. Analysts often use the four-fifths rule as a benchmark: if the selection rate for one group is less than 80 percent of the rate for the highest-selected group, that gap signals potential disparate impact worth examining.
Before S8338, disparate impact claims under the NYSHRL rested on case law rather than an express statutory standard. The amendment formally codifies the theory, resolving ambiguity and aligning the NYSHRL with federal law and the New York City Human Rights Law. It also closes a geographic gap that Local Law 144 leaves open: because the NYSHRL applies everywhere in New York State, workers outside the five boroughs now have a clearer path to challenging AI hiring tools regardless of whether their employer was ever subject to the city’s bias-audit requirement.
Federal law offers parallel protections. Title VII of the Civil Rights Act, the Age Discrimination in Employment Act (ADEA), and the Americans with Disabilities Act (ADA) all provide grounds for challenging employment decisions with discriminatory effects. The Equal Employment Opportunity Commission has also issued algorithmic discrimination guidance affirming that AI-based hiring tools are subject to the same anti-discrimination standards as any other employment practice. New York’s amended NYSHRL now provides a state-level counterpart with its own enforcement mechanisms and remedies.
Recognizing the Signs of Algorithmic Screening
Most applicants rejected by an AEDT don’t know it happened. The rejection looks like any other. There are patterns worth watching for, though none is definitive on its own:
- Immediate rejection: A rejection arriving within minutes or a few hours of submitting an application, before any human could have reviewed it.
- No human contact: No recruiter reaches out at any stage, and the rejection comes from a noreply address or automated system.
- Generic language despite a strong match: The rejection uses form language even though your qualifications closely match the posted requirements.
- Video interview with no follow-up: You completed an asynchronous video interview through a platform known to score facial expressions or tone, followed by an automated rejection.
Workers who suspect an AEDT was involved can send a written request to the employer asking whether an automated decision tool was used and what data it evaluated. Current law doesn’t guarantee access to your individual score, but the employer’s response (or non-response) can be relevant if a claim proceeds. Preserving documentation from the start builds the evidentiary record a discrimination claim depends on: the original job posting, application confirmation emails, timestamps on any automated messages, and your full correspondence with the employer.
Filing Deadlines That Can Make or Break Your Claim
Deadlines for AI hiring discrimination claims in New York vary significantly depending on which forum you use, and one of the most important changed in 2024. Missing the wrong one means losing the right to pursue a claim entirely.
NYC Commission on Human Rights
For most discrimination claims under the New York City Human Rights Law, you have one year from the discriminatory act to file a complaint with the NYC Commission on Human Rights. For gender-based harassment claims, that window extends to three years. If you choose to file a civil action directly in court under the NYCHRL rather than going through the Commission, the statute of limitations is also three years.
NYS Division of Human Rights
For discrimination incidents occurring on or after February 15, 2024, you now have three years to file an administrative complaint with the New York State Division of Human Rights, a significant change from the previous one-year deadline. If your incident occurred before February 15, 2024, the old one-year deadline still applies.
Federal EEOC
Claims under Title VII, the ADEA, or the ADA require filing a charge with the Equal Employment Opportunity Commission before you can bring a federal lawsuit. In New York, you have 300 days from the discriminatory act to file that charge. Missing this deadline generally bars federal claims regardless of their merit.
These deadlines run from the discriminatory act itself. In an AI hiring context, that is typically the date of the automated rejection. Waiting to see whether a reapplication or appeal goes anywhere can quietly consume time you can’t recover.
What to Do If You Were Screened Out Unfairly
Acting early matters for more than one reason. Evidence in AI hiring cases is often held entirely by the employer: the bias audit data, the algorithm’s scoring criteria, the training data, and your individual evaluation record all sit on the employer’s side of the dispute. The Comptroller’s audit findings suggest that what appears on an employer’s public website may not reflect what the tool actually does. Building a claim often means obtaining that information through formal legal process before it changes or disappears.
Forum choice matters too. A worker who misses the 300-day federal deadline but acts within three years still has viable state claims. A worker who waits more than three years may lose everything. And the forum affects not just timing but available remedies, how the case is investigated, and whether it can ultimately be brought in court.
If you were rejected from a New York job and suspect an algorithm played a role, our attorneys at Schwartz Perry & Heller LLP can evaluate whether that rejection crossed the legal line. Reach us at (646) 490-0221.