New Jersey AI employment law 2026: employer compliance guide

TL;DR: New Jersey has not enacted a standalone AI employment disclosure law. Two existing frameworks already create real obligations: the NJLAD prohibits discriminatory AI hiring outcomes and the Division on Civil Rights' Disparate Impact Rules (effective December 15, 2025) specifically address automated tools. Proposed bills A3854 and A3855 would add disclosure, opt-out, and bias audit requirements, treat them as planning targets.

New Jersey has not passed a standalone AI employment disclosure law, but the state already imposes meaningful obligations on employers using AI in hiring. The New Jersey Law Against Discrimination (NJLAD) and the Division on Civil Rights' sweeping Disparate Impact Rules, adopted December 15, 2025 as the most comprehensive state-level disparate impact rules in the country, prohibit automated hiring tools that produce discriminatory outcomes. Separately, proposed bills A3854 and A3855 would add explicit disclosure requirements, opt-out mechanisms, and independent bias audit mandates. This guide covers what NJ employers must do under current law and how to prepare for where proposed legislation is clearly headed.

What NJ law currently requires

NJ Law Against Discrimination (NJLAD)

The NJLAD is New Jersey's primary anti-discrimination statute and applies directly to employers using AI in hiring. The NJLAD prohibits employment practices that cause disparate impact across protected classes, regardless of intent and regardless of whether a human or an automated system produced the outcome. "The algorithm did it" is not a NJLAD defense.

DCR Disparate Impact Rules (effective December 15, 2025)

The Division on Civil Rights adopted rules on December 15, 2025 that codify how disparate impact applies to AI and automated hiring tools. These are the most comprehensive state-level disparate impact rules in the country. Key employer obligations:

• Automated tools face full disparate impact analysis. Resume screening algorithms, video interview analysis tools, online application filters, and AI-generated hiring recommendations must not produce unjustified disparate impact across race, sex, national origin, disability, or other protected classes.
• Vendor liability does not shield employers. If a vendor's AI product causes disparate impact, the employer remains liable. Employers must take reasonable steps to ensure vendor tools comply, including requesting demographic impact data and bias testing results.
• Burden-shifting applies. If a tool produces disparate impact, the employer must show the practice is necessary to achieve a substantial, legitimate, nondiscriminatory interest and no less discriminatory alternative exists.

Proposed legislation: what bills A3854 and A3855 would add

Neither bill has been signed into law as of mid-2026. Treat them as planning targets:

• A3854 would require explicit disclosure to applicants before automated tools are used, a real opt-out mechanism for human review, and a prohibition on AI being the sole determinant of a hiring decision.
• A3855 would require employers to commission independent annual bias audits of AI hiring tools, modeled on NYC Local Law 144.

The compliance steps below reflect both current NJLAD/DCR obligations and preparation for proposed requirements. Items labeled "current law" are required now; items labeled "proposed" are building infrastructure for likely future requirements.

Who is covered

Proposed bill A3854 would apply to employers that use automated employment decision tools in connection with hiring, promotion, or any employment decision affecting individuals employed in or applying for jobs in New Jersey. Coverage would not be limited to large employers. Under the proposed definition, a five-person company using an AI resume screener to filter applicants for a New Jersey position would be covered.

Staffing agencies and HR technology vendors that deploy these tools on behalf of employers are also in scope. Employers should audit vendor relationships and ensure contractual obligations for disclosure and opt-out support are clearly allocated.

The law covers decisions related to:

• Initial hiring and candidate screening
• Internal promotions and role changes
• Performance evaluations that drive employment outcomes
• Layoff or reduction-in-force scoring

What counts as an automated employment decision tool

The definition in proposed bill A3854 tracks language now familiar from NYC Local Law 144 and similar state proposals: a computational process derived from machine learning, statistical modeling, data analytics, or artificial intelligence that produces a simplified output (score, classification, rank, or recommendation) used to substantially assist or replace discretionary human judgment.

Practical examples of tools that are almost certainly covered:

• Resume screening platforms that score and rank candidates
• Video interview analysis software that evaluates speech patterns, facial expressions, or word choice
• Personality and cognitive assessment platforms that generate a hiring recommendation
• Predictive attrition or fit models used in promotion decisions

Practical examples of tools that are less likely to be covered:

• Applicant tracking systems used only for scheduling and status tracking, with no scoring or ranking output
• Simple keyword filters where a human reviews all results without reference to an AI-generated score
• Background check providers that report factual records without generating a hiring recommendation

When in doubt, apply the disclosure requirement. Disclosing when you did not need to is far less risky than failing to disclose when you did.

How NJ compares to NYC Local Law 144

Many employers with New Jersey operations are already familiar with NYC Local Law 144, which has been enforced since July 2023. The two laws share the same philosophical foundation but differ in one important way.

The absence of an independent bias audit requirement is the most significant difference. New Jersey employers do not yet need to hire a third-party auditor to test their AI tools for disparate impact across demographic groups. However, that gap may close. The NJ legislature has active proposals that would add an audit requirement modeled on the NYC approach.

For employers already running bias audits for NYC compliance, extending that process to cover New Jersey operations is low additional cost and reduces exposure if the audit requirement is added.

Sample disclosure language

The following language can be adapted for use in job postings, application portals, and offer letter cover pages. This is a starting point, not legal advice. Have your employment counsel review final language.

AI Tool Disclosure Notice

[Company Name] uses automated tools in its hiring process. These tools may analyze your resume, application materials, or interview responses to generate a score or recommendation that assists our hiring team. The tools do not make final hiring decisions; all decisions are reviewed by a human member of our team.

If you prefer to have your application evaluated without the use of automated scoring tools, you may request a manual review by contacting [HR contact name] at [email address] or [phone number]. We will acknowledge your request within [X] business days and confirm that your application has been reviewed by a member of our recruiting team.

For questions about the specific tools we use, please contact [HR contact].

For internal promotions and performance-related decisions, a similar notice should be sent to the affected employee before the tool is used, not after the decision is made.

Employer compliance checklist

Work through this checklist before the law's effective date. Assign an owner and target date for each item.

Step 1: Inventory your AI tools

• List every software tool used in hiring, screening, and promotion decisions affecting New Jersey employees or applicants
• For each tool, determine whether it produces a score, ranking, classification, or recommendation used in employment decisions
• Flag tools that meet the definition of an automated employment decision tool

Step 2: Update job postings and application materials

• Add AI disclosure language to all job postings that involve use of a covered tool
• Update the application portal to present the opt-out request mechanism before the applicant submits their application
• Archive the old versions of postings to document when the update was made

Step 3: Create and document the opt-out process

• Designate an HR owner for opt-out requests by job category or region
• Set a written response time (recommend 3-5 business days)
• Create a tracking log so you can demonstrate that opt-out requests were honored
• Test the process with a mock request before going live

Step 4: Audit your decision workflows

• Review every hiring and promotion workflow that uses a covered AI tool
• Confirm that a human reviewer sees the AI output and has the authority and practical ability to override it
• Document the human review step in your hiring process documentation

Step 5: Train HR and recruiting teams

• Brief all recruiters on what the disclosure requirement means and when it applies
• Train HR business partners on the opt-out process, including how to handle requests quickly and how to document them
• Update onboarding materials for new HR staff

Step 6: Review vendor contracts

• Identify all third-party HR tech vendors that provide covered tools
• Add data processing addenda and contractual representations confirming vendors can support your disclosure and opt-out obligations
• Require vendors to notify you if they make material changes to their AI scoring logic

Step 7: Monitor legislative developments

• Track NJ legislative activity for proposed bias audit requirements, private right of action provisions, and expanded scope
• Set a calendar reminder to review compliance at least quarterly in 2026-2027

What is coming in New Jersey

New Jersey has been one of the more active states on AI employment regulation. As of mid-2026, the legislative pipeline includes:

• A3854 (AEDT disclosure and opt-out): Would require explicit notice to applicants before automated tools are used, a real opt-out mechanism for human review, and prohibition on AI as the sole determinant. In committee.
• A3855 (independent bias audits): Would require annual third-party bias audits of AI hiring tools, similar to NYC Local Law 144. Status was "Introduced" as of 2024; monitoring continued.
• Private right of action: Some proposals would give individual applicants the ability to sue directly rather than routing all enforcement through the DCR.
• Expanded scope: Proposals would extend coverage to performance management, automated scheduling, and AI-assisted termination decisions.

The DCR Disparate Impact Rules enacted in December 2025 already give the Division active enforcement tools against AI-driven discrimination without waiting for additional legislation. Employers should not treat the absence of a signed disclosure law as permission to skip assessment. Build your infrastructure now.

Related reading

For the broader state-by-state picture, see AI hiring tool compliance US state laws. For the NYC requirements that predate NJ, see NYC Local Law 144 AI bias audit employer guide. The EEOC AI hiring guidance 2026 employer checklist covers federal overlay requirements that apply regardless of state law. See the AI regulation deadline calendar 2026 for effective dates across all major US AI laws. The AI governance checklist 2026 covers broader governance requirements beyond employment. For FCRA obligations that often accompany AI hiring tool use, see FCRA AI hiring disclosure requirements 2026. For Colorado's upcoming framework, see Colorado AI Act SB 189 2027 employer guide. For the enacted replacement law under SB 26-189, see Colorado AI Act SB 26-189: employer compliance guide 2027. For Illinois requirements, see Illinois AI employment disclosure law 2026. For a recent vendor liability example, see Workday AI lawsuit HR screening checklist. For the broader multi-state picture including Utah's AI policy framework, see Utah AI policy act compliance 2026.

• AI Bias Audit Requirements by State (2026): NYC, Colorado, Minnesota, Ne