New Jersey AI Laws 2026: Complete Guide for Businesses

New Jersey has emerged as one of the most active state legislatures on AI regulation in 2026, with 137 published AI bills across consumer protection, biometric surveillance, algorithmic pricing, deepfakes, and AI governance. While the state has not yet enacted a comprehensive AI statute, a cluster of targeted bills — several already advanced out of committee — signals a clear regulatory direction: consumer-first, enforcement-through-existing-fraud-law, and the governor publicly committed to signing key measures. For businesses operating in New Jersey, the legislative pipeline is not a future concern; it is a present compliance risk.

New Jersey AI Legislation at a Glance

As of October 2026, New Jersey's AI legislative landscape includes 137 published AI bills: 25 introduced and 96 in committee. The five most active topic areas by bill count are AI Deepfakes (38 bills), AI Transparency (23 bills), Comprehensive AI (16 bills), AI in Government (14 bills), AI in Employment (10 bills). For the full bill index, visit our New Jersey AI laws page.

Biometric Surveillance: A3929

Assembly Bill 3929 prohibits business entities from using a biometric surveillance system on consumers at their physical premises unless two conditions are both satisfied: (1) the business provides clear and conspicuous notice to consumers of the system's use, and (2) the biometric surveillance system is used for a lawful purpose. The notice requirement can be met by posting a sign in a conspicuous location at the perimeter of any area where the system operates.

Heightened Duties When Denying Access

A3929 imposes a more demanding standard when biometric data drives an adverse access decision. If a business uses information obtained through biometric surveillance to deny a consumer entry to its premises or to remove a consumer from its premises, the business must provide that consumer with a detailed written explanation of its actions and the criteria applied. This creates a de facto adverse-action notice obligation analogous to those found in credit and employment screening statutes.

Data Commercialization Prohibition

The bill prohibits business entities from selling, leasing, trading, sharing, or otherwise profiting from information obtained through their biometric surveillance systems. This prohibition broadly tracks the architecture of the Illinois Biometric Information Privacy Act (BIPA) but applies specifically in the consumer-facing commercial context.

Cure Period

A business will not be liable for a first violation if it demonstrates compliance within 30 days of notice of the violation. Subsequent violations do not benefit from this cure provision. A3929 was reported favorably out of the Assembly Consumer Affairs Committee on March 19, 2026. See our facial recognition laws tracker for related measures across other states.

Facial Recognition in Retail: S3268

Senate Bill 3268 takes a direct prohibitory approach: it bars the use of facial recognition technology on consumers, with a narrow exception for legitimate safety purposes. The bill supplements P.L.1960, c.39 (the Consumer Fraud Act), meaning violations would be actionable under the Attorney General's existing enforcement authority and the CFA's treble-damages and attorney's fees provisions. Unlike A3929, S3268 contains no notice-and-use framework — use for non-safety commercial purposes would be categorically prohibited. The bill takes effect immediately upon enactment. Track status on our facial recognition page.

Algorithmic Pricing: A Package of Bills

New Jersey has introduced more algorithmic pricing legislation than any other single topic cluster in its 2026 session. the governor has publicly stated she will sign the primary bills if they reach her desk, making this the area of most imminent legal risk for retailers and consumer-facing businesses.

A4742 — Fair Pricing and Transparency Act

Assembly Bill 4742, introduced March 19, 2026, by legislators in District 33, targets retail food stores. Its core prohibitions: a retail food store may not use (1) surveillance-based price discrimination, (2) dynamic pricing, or (3) personalized pricing to modify the sale price of groceries or other foodstuffs sold to consumers. "Personalized pricing" is defined as pricing determined or modified, in whole or in part, through an automated system, AI, a machine learning model, or an algorithm that relies on consumer data or inferred data — including browsing history, shopping history, geolocation, biometric information, and demographic inferences such as income, race, or disability status.

The bill carves out discounts and loyalty program benefits that are uniformly offered to all consumers who meet disclosed criteria, provided that personally identifiable information collected to administer those programs is not used to personalize prices. A4742 also requires the Division of Consumer Affairs to conduct a study, within 12 months of the effective date, on electronic shelf label systems and their impact on pricing transparency and employee job security. Violations are unlawful practices under the Consumer Fraud Act. The bill carries an effective date of October 1, 2026.

A4685 — Algorithmic Device Pricing Ban

Assembly Bill 4685 is broader in scope: it prohibits the use, sale, and provision of access to algorithmic devices that set the sale price of goods or services — not limited to food retail. This bill would reach any business entity in New Jersey using an algorithm-driven pricing engine, including software vendors who supply such tools. If enacted, it would take effect the first day of the seventh month after enactment. See our algorithmic pricing laws guide for the national landscape.

A4085 / S3612 — Fair Price Protection Act (enacted for groceries)

Senate Bill 3612 was introduced as a broad ban on personalized algorithmic and surveillance pricing of merchandise and services under the Consumer Fraud Act. The version that became law is its Assembly counterpart, A4085, which narrowed the ban to groceries and other foodstuffs. Signed July 23, 2026 as P.L.2026, c.55, the Fair Price Protection Act takes effect Aug. 1, 2027, with exceptions for cost-based changes, publicly available discounts and opt-in loyalty programs, and a one-year moratorium on new electronic shelf labels starting Feb. 1, 2027. Violations are unlawful practices under the Consumer Fraud Act. See our Surveillance Pricing Tracker.

S3732 — Grocery Dynamic Pricing Ban

Senate Bill 3732 focuses specifically on retail food stores and third-party grocery delivery platforms, prohibiting the use of dynamic pricing, surveillance pricing, or personalized algorithmic pricing when selling groceries — whether in-store or online. Like S3612, it enforces through the Consumer Fraud Act. Loyalty program discounts based on prior purchase history are expressly permitted. It remains in committee; the enacted A4085 now covers grocery pricing.

A3989 — Rent Algorithmic Pricing Restriction

Assembly Bill 3989 extends the algorithmic pricing prohibition to the residential rental market, restricting the use and sale of algorithmic devices that set the rent price or determine occupancy of residential property. The bill takes effect the first day of the second month following enactment.

Deepfake and Synthetic Media: S3303

Senate Bill 3303, introduced February 2, 2026 by a District 40 senator, amends New Jersey's existing voyeurism and nonconsensual disclosure criminal statutes to expressly cover AI-generated deepfake images. Under the bill, a "deepfake" is defined as an image, video, or audio recording of speech or conduct that appears to a reasonable person to realistically depict the speech or conduct of a person who did not in fact engage in that speech or conduct, with that appearance created or enhanced by artificial intelligence, machine learning, or similar technology.

Criminal Exposure

• Creating a nonconsensual pornographic deepfake image is a crime of the third degree (punishable by three to five years imprisonment, a fine of up to $15,000, or both).
• Disclosing such an image is also a crime of the third degree.
• Disclosing a deepfake image created in violation of the statute, with reckless disregard, is a crime of the fourth degree.
• Threatening to disclose a nonconsensual pornographic deepfake image constitutes a criminal threat under the existing extortion framework.

This bill complements but does not replace New Jersey's existing non-AI nonconsensual intimate image (NCII) laws. See our AI deepfakes by state tracker for the full national picture.

Child Endangerment and AI: A2767

Assembly Bill 2767, introduced January 13, 2026 and sponsored by a District 37 legislator, modifies New Jersey's child endangerment statute to include AI technology. The bill makes it a crime of the fourth degree for any person to engage in the control, possession, manipulation, use, dissemination, sale, or promotion of any automated interactive computer technology that is capable of performing or simulating sexual conduct. A crime of the fourth degree is punishable by up to 18 months imprisonment and a fine of up to $10,000.

The bill provides a safe harbor: a person is not in violation if the automated interactive computer technology has safety features designed to stop a chatbot from simulating sexual conduct upon detecting that a user is a child. The bill was referred to the Assembly Public Safety and Preparedness Committee upon introduction. The effective date in the database is January 1, 2026 — practitioners should track its enactment status closely.

AI Companion Disclosure: A4732

Assembly Bill 4732 requires operators of AI companion systems to provide clear and conspicuous notification — verbally or in writing — at the beginning of every interaction that the user is not communicating with a human. For interactions that continue for an extended period, the notification must be repeated at least every three hours. The bill was reported favorably out of the Assembly Science, Innovation and Technology Committee on March 16, 2026 and referred to the Assembly Children, Families, and Food Security Committee.

The bill defines "AI companion" with specificity: a system using generative AI designed to simulate a sustained human or human-like relationship by (1) retaining information from prior interactions to personalize ongoing engagement, (2) asking unprompted emotion-based questions beyond direct responses to user prompts, and (3) sustaining ongoing dialogue about matters personal to the user. Civil penalty: $15,000 per violation. This definition is narrower than some comparable state bills and focuses on relationship-simulation features rather than all AI chatbot applications.

AI Transparency in Commerce: A4730 and A4733

Assembly Bill 4730 requires any person or entity to notify consumers when they are communicating with generative AI in the context of trade or commerce — a broad disclosure obligation that would apply to AI-driven customer service, sales interactions, and similar commercial contexts. The bill was voted out of the Assembly Science, Innovation and Technology Committee in March 2026 and referred to the Assembly Consumer Affairs Committee.

Assembly Bill 4733 prohibits advertising generative AI as able to practice a regulated profession or occupation. This bill directly addresses AI tools marketed with claims that they can perform legal, medical, financial, or other licensed professional services. It passed the Assembly Science, Innovation and Technology Committee and was referred to the Assembly Regulated Professions Committee. A related bill, S4463/A5603 (introduced May 2025), prohibits AI systems from advertising themselves as licensed mental health providers.

AI in Schools: A1323

Assembly Bill 1323, sponsored by a legislator from the Atlantic, Camden, and Gloucester county district, passed the full General Assembly in February 2026. It requires boards of education in school districts, charter schools, and renaissance school projects to adopt formal policies governing the use of AI-enabled surveillance systems — including AI-enhanced video cameras, x-ray machines, and body scanners — if such systems are deployed. Required policy elements include: what systems are in place, what information is collected, how it is used and shared, and how families can access the information. Schools must also post signage in prominent locations notifying the public when AI surveillance is in use, and distribute the policy to parents and guardians.

AI Governance and Data Centers

Several NJ bills address AI governance at the state level and the energy demands of AI infrastructure:

• A4821 / S3742 (introduced September–October 2024) would require the Office of Information Technology (OIT) to establish AI safety requirements and mandate that AI companies conduct safety tests and report results to the OIT.
• S3357 (introduced June 2024) would establish the New Jersey Artificial Intelligence Advisory Council — a seven-member body including the Chief Technology Officer, legislative appointees, and gubernatorial appointees with AI expertise — to assess AI ethics codes for state agencies, review automated decision systems, and recommend procurement guidelines.
• A4769 / S3639 would require state agencies including BPU, DCA, and DEP to establish expedited approval and permitting procedures for AI data centers powered by small modular nuclear reactors.
• S680 would require AI data centers and cryptocurrency mining facilities in New Jersey to derive all electricity from new clean energy sources.
• A4710 (Clean Energy AI Incentivization Act) would direct the Board of Public Utilities to incentivize AI data centers to utilize self-sufficient clean energy sources.
• A2616 would require OIT to develop a New Jersey generative AI program available to the public, subject the program to annual safety testing, and coordinate AI education courses with county governments, with an appropriation of $1 million.

Track all AI governance bills on our AI in Government topic page.

Key Bills Summary Table

Compliance Priorities by Company Type

What Distinguishes New Jersey's Approach

Several features distinguish New Jersey's AI regulatory strategy from other active states:

• Consumer Fraud Act as the enforcement vehicle. Nearly all of New Jersey's AI consumer protection bills enforce through the existing Consumer Fraud Act (P.L.1960, c.39), giving the Attorney General broad enforcement authority and plaintiffs access to treble damages, attorney's fees, and punitive damages without requiring a new enforcement agency or litigation framework.
• Algorithmic pricing as a legislative priority. New Jersey has introduced more algorithmic pricing bills per capita than most other states, and the governor's public commitment to signing the primary grocery pricing bills gives this cluster unusual urgency.
• Targeted criminalization rather than comprehensive regulation. Unlike states pursuing omnibus AI governance frameworks, New Jersey is adding AI-specific provisions to existing criminal statutes (deepfakes, child endangerment) rather than enacting a standalone AI law.
• Federal preemption risk. Legal analysts have flagged that pending federal executive action seeking to preempt state AI regulation may create conflicts with several of these bills. Businesses should monitor both state enactment and any federal preemption developments in parallel.

For a complete index of New Jersey AI legislation, visit our New Jersey AI laws page. For cross-state comparison, see our trackers for Algorithmic Pricing, Facial Recognition, and AI Deepfakes.

This article is for informational purposes only and does not constitute legal advice. Bill provisions, statuses, and effective dates are subject to change. Consult a qualified attorney for guidance specific to your situation.

Sources: NJ Legislature (pub.njleg.state.nj.us); LegiScan (legiscan.com); Crowell & Moring client alert (crowell.com, March 17, 2026); Troutman Privacy + Cyber + AI update (troutmanprivacy.com, March 23, 2026); Pashman Stein Walder Hayden P.C. (pashmanstein.com, December 17, 2025); NJ Assembly Democrats press release (assemblydems.com, February 24, 2026); New Jersey Globe (newjerseyglobe.com, March 29, 2026).