AI Rejected Your Job Application in NJ: How Algorithmic Hiring Tools Can Violate Anti-Discrimination Law

AI-driven hiring tools are increasingly used in New Jersey job applications. Employers rely on automated systems for efficiency. But those systems reflect biased data and apply criteria that disproportionately impact protected classes. Even when it looks neutral, it still leads to unequal access to opportunities.

In many cases our legal team at Brandon J. Broderick reviews, applicants are rejected without clear explanations. When decisions are influenced by algorithms, the focus shifts to how those systems are trained and applied in practice.

When hiring algorithms screen out applicants based on protected characteristics, this leads to discriminatory outcomes under both New Jersey and federal law.

In this guide, we discuss how AI hiring tools operate, what patterns point to legal concerns, how they can help applicants recognize their rights, and when it is time to consult an employment lawyer in New Jersey.

AI Screening in New Jersey Hiring and What It Means for Your Application

Employers across New Jersey increasingly rely on automated systems to manage large volumes of applicants. These systems are designed to screen and narrow the field before a recruiter or hiring manager steps in. 

Recent data shows that 63% of employers now use at least one AI-powered tool in recruiting or hiring decisions. Nearly half (47%) also admit that these systems can produce unfair or biased outcomes.

Here is how that typically works.

Resume parsing

When you upload your résumé, the system breaks the document into data points: job titles, dates, skills, and education. It then converts that information into structured fields inside a database. 

If the formatting is unusual, or key information isn’t in expected locations, the system may not “see” these parts of your experience.

This often affects work that’s not easily reflected in standard formats. If workers were paid partially in cash, as sometimes happens with misclassified contractors, it would not appear in a way that the system can evaluate.

Keyword filtering

Keyword matching is one of the most common tools used in applicant screening systems.

Employers load the system with preferred terms tied to the job description. The tool scans résumés for those words. If your experience is described differently, your score drops. For example, a résumé that says “client relations” instead of “customer success” is ranked lower.

Video interview scoring

Recorded video interviews have become a common part of the hiring process. The program can analyze speech patterns, word choice, tone, or pace of response. Certain accents or speech patterns are often scored as less favorable by biased AI screening, sometimes reflecting underlying racial stereotypes.

This generates scores based on subjective traits like confidence, communication skills, or “fit.” 

Personality and behavioral assessments

In online questionnaires, applicants answer timed questions designed to predict work style or reliability. The results are translated into numerical rankings. If the score falls outside a preset range, the candidate is automatically removed.

In recent years, our legal team at Brandon J. Broderick has seen this play out across industries. The result is often a silent rejection, with no clear indication that automation was involved.

The decision looks automated, but it is still considered the employer’s decision. The technology changes how the decision is delivered, but it doesn’t remove legal accountability.

Where Algorithmic Bias Comes From in NJ Employment Decisions

“AI bias” usually comes from how the system is built: what data it learns from and what signals it is told to prioritize.

When a company historically favored certain schools or job histories, the system views those examples as indicators of success. If bias existed before, the system scales it. When prior roles offered limited access to training opportunities, the limitation followed the applicant.

Even without directly using protected traits, systems rely on inputs that closely track them.

• Zip codes can reflect race or income levels
• Employment gaps can relate to disability, pregnancy, or caregiving
• School names that signal socioeconomic background

Some systems score candidates based on response speed, while others favor certain job titles. These systems ignore context. That includes nontraditional careers, or leave taken under New Jersey’s domestic violence law that a human reviewer might recognize.

The system looks neutral, but the outcomes are biased. Candidates ranked higher are more likely to be hired. The data feeds back into the system, strengthening the same preferences.

NJLAD Guidance on AI Hiring and Disparate Impact

Under the New Jersey Law Against Discrimination, hiring decisions are judged by their impact. An employer doesn’t need to mean to discriminate. If a system produces unequal results, it’s biased. 

Disparate treatment involves unequal treatment tied to a protected characteristic.

• A tool is set up to treat applicants differently based on a protected trait, either directly or through clearly linked inputs
• Screening rules are applied unevenly — for example, certain applicants are pushed through extra filters or assessments while others aren’t
• “Selective use” of a tool becomes part of the problem — using stricter or additional automated screening only for certain groups

The system creates a difference in treatment from the start. More often, the problem shows up as disparate impact to produce uneven results.

• The same tool is applied to everyone, using the same criteria
• There is no obvious reference to a protected trait in the design
• But over time, one group is screened out at a higher rate than others

In 2025, New Jersey adopted formal disparate impact regulations. A resume screener can appear neutral and still exclude certain applicants. 

The Equal Employment Opportunity Commission treats algorithmic systems as selection procedures. They can be evaluated the same way as traditional tests. Looking at the outcomes:

• Comparing selection rates across different groups
• Identifying if one group is consistently filtered out more often than others
• Asking whether the criteria being used are actually necessary for the job

This kind of analysis focuses on what the tool does in practice.

In December, the Attorney General and DCR issued guidance addressing algorithmic decision-making. Employers cannot rely on the argument that a system is neutral or widely used. Saying “we didn’t know” or “the vendor built it” doesn’t resolve the problem under NJLAD.

AI systems are often used beyond initial hiring. Screening systems are also used for promotions, discipline, layoffs, and severance offers, where historical data still reflects bias.

Disability Rights and Automated Hiring Tools in New Jersey

Employers are not targeting disability, but automated systems often favor uniform behavior. Timed tests or typing speed requirements can disadvantage applicants with disabilities. This happens when systems don’t account for accommodations like assistive technology.

Some tools are built around fixed assumptions about how a “qualified” candidate should perform during the hiring process:

• Screening tools that reject applicants based on gaps in work history
• Video interview software scores eye contact, tone, or facial expressions in ways that penalize neurological differences
• Chatbots or rigid application flows may not allow alternative ways to complete required steps

In each of these situations, the problem is not the applicant’s ability to do the job, but whether the hiring process gives them a fair chance to show it. If an automated tool interferes with an applicant’s ability to request or use a reasonable accommodation, it raises legal concerns. 

The EEOC and U.S. Department of Justice warn that employers remain responsible when hiring tools screen out individuals with disabilities. The focus stays on access and fairness, regardless of who built the system.

An automated rejection can feel like a simple mismatch. But there are certain red flags:

• No option to request accommodations during online assessments
• Instructions that assume a specific pace or communication style
• Immediate rejection after completing a standardized test or video interview
• Lack of any alternative process when a tool doesn’t fit the applicant’s needs

Under the Americans with Disabilities Act, employers must provide reasonable accommodations during the hiring process.

What to Do If You Suspect AI Hiring Discrimination in New Jersey

This is where our specialists often tell clients to begin when building a strong claim. Start with what you already have:

• Save the job posting and note when you applied and when the rejection came through
• Keep screenshots of any assessments, video interviews, or questionnaires
• Hold on to confirmation emails and any communication about the process
• If you asked for an accommodation, keep a record of that request and the response

Next, ask a few direct questions. Keep them in writing:

• Was an automated tool used at any stage of the hiring process
• Are accommodations available for assessments or screening tools
• Whether a human review of your application is possible

If you are dealing with potential hiring discrimination, seeking legal guidance can help you understand your rights and what steps to take next.

When Biased Hiring Algorithms Cross the Legal Line

Automated hiring tools can speed up decisions and help manage large applicant pools. 

But the use of technology doesn’t change the standard. New Jersey still requires fair access to jobs and prohibits screening methods that produce biased results.

If you believe an automated process may have crossed that line, contact us today for a free consultation.