New Jersey · Senate Bill · 221st Legislature
SB3742
New Jersey Senate Bill 3742 — An Act concerning artificial intelligence safety tests and supplementing Title 52 of the Revised Statutes

Status ● Introduced Effective N/A Passage Likelihood M

WHAT THIS BILL REGULATES · 3 REQUIREMENT TYPES

How Is This Bill Enforced

Enforcement Authority
Office of Information Technology is responsible for establishing safety test requirements, reviewing annual reports, and adopting rules and regulations. No private right of action is created. No penalty or enforcement mechanism is specified in the bill text.
Private Right of Action
No private right of action. Enforcement is exclusive to the designated authority.
Penalties
The bill does not specify any penalties, damages, or remedies for non-compliance.

What This Bill Requires

Verbatim statutory text on the left; plain-language analysis and a per-section checklist on the right. Numbered markers cross-link to the matching checklist row.

Statutory Text
Analysis & Obligations
Section 1(a)
Definitions

As used in P.L. , c. (C. ) (pending before the Legislature as this bill): "Artificial intelligenceArtificial intelligence"Artificial intelligence" means an artificial intelligence system: a. that performs tasks under varying and unpredictable circumstances without significant human oversight or that can learn from experience and improve performance when exposed to data sets; b. developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action; c. designed to think or act like a human, including an artificial intelligence system that has cognitive architectures and neural networks; d. that uses a set of techniques, including machine learning, that is designed to approximate a cognitive task; or e. designed to act rationally, including an intelligent software agent or embodied robot that achieves goals using perception, planning, reasoning, learning, communicating, decision making, and acting.Section 1(a)" means an artificial intelligenceArtificial intelligence"Artificial intelligence" means an artificial intelligence system: a. that performs tasks under varying and unpredictable circumstances without significant human oversight or that can learn from experience and improve performance when exposed to data sets; b. developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action; c. designed to think or act like a human, including an artificial intelligence system that has cognitive architectures and neural networks; d. that uses a set of techniques, including machine learning, that is designed to approximate a cognitive task; or e. designed to act rationally, including an intelligent software agent or embodied robot that achieves goals using perception, planning, reasoning, learning, communicating, decision making, and acting.Section 1(a) system: a. that performs tasks under varying and unpredictable circumstances without significant human oversight or that can learn from experience and improve performance when exposed to data sets; b. developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action; c. designed to think or act like a human, including an artificial intelligenceArtificial intelligence"Artificial intelligence" means an artificial intelligence system: a. that performs tasks under varying and unpredictable circumstances without significant human oversight or that can learn from experience and improve performance when exposed to data sets; b. developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action; c. designed to think or act like a human, including an artificial intelligence system that has cognitive architectures and neural networks; d. that uses a set of techniques, including machine learning, that is designed to approximate a cognitive task; or e. designed to act rationally, including an intelligent software agent or embodied robot that achieves goals using perception, planning, reasoning, learning, communicating, decision making, and acting.Section 1(a) system that has cognitive architectures and neural networks; d. that uses a set of techniques, including machine learning, that is designed to approximate a cognitive task; or e. designed to act rationally, including an intelligent software agent or embodied robot that achieves goals using perception, planning, reasoning, learning, communicating, decision making, and acting. "Artificial intelligence companyArtificial intelligence company"Artificial intelligence company" means a private entity or public agency that sells, develops, deploys, uses, or offers for sale in this State artificial intelligence technology.Section 1(a)" means a private entity or public agency that sells, develops, deploys, uses, or offers for sale in this State artificial intelligence technologyArtificial intelligence technology"Artificial intelligence technology" means a product or service that primarily uses artificial intelligence to perform its intended function.Section 1(a). "Artificial intelligence technologyArtificial intelligence technology"Artificial intelligence technology" means a product or service that primarily uses artificial intelligence to perform its intended function.Section 1(a)" means a product or service that primarily uses artificial intelligenceArtificial intelligence"Artificial intelligence" means an artificial intelligence system: a. that performs tasks under varying and unpredictable circumstances without significant human oversight or that can learn from experience and improve performance when exposed to data sets; b. developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action; c. designed to think or act like a human, including an artificial intelligence system that has cognitive architectures and neural networks; d. that uses a set of techniques, including machine learning, that is designed to approximate a cognitive task; or e. designed to act rationally, including an intelligent software agent or embodied robot that achieves goals using perception, planning, reasoning, learning, communicating, decision making, and acting.Section 1(a) to perform its intended function. "Safety testSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a)" means a protocol for assessing artificial intelligence technologyArtificial intelligence technology"Artificial intelligence technology" means a product or service that primarily uses artificial intelligence to perform its intended function.Section 1(a) for potential biases, inaccuracies, and cybersecurity threats.

Section 1(a) establishes the four defined terms that frame the bill's scope. Artificial intelligence company is defined expansively to include any private entity or public agency that sells, develops, deploys, uses, or offers AI technology for sale in New Jersey — capturing not only developers and vendors but also end-user organizations. Artificial intelligence technology is any product or service that primarily uses AI to perform its intended function, which introduces a threshold question about what constitutes "primarily." The definition of artificial intelligence itself is a broad disjunctive list drawn from multiple AI conceptual frameworks.

Section 1(b)
OIT safety test standard-setting and report review duties
Government

(1)(a)–(c) 1 The Office of Information Technology shall: (1) establish minimum requirements for an artificial intelligenceArtificial intelligence"Artificial intelligence" means an artificial intelligence system: a. that performs tasks under varying and unpredictable circumstances without significant human oversight or that can learn from experience and improve performance when exposed to data sets; b. developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action; c. designed to think or act like a human, including an artificial intelligence system that has cognitive architectures and neural networks; d. that uses a set of techniques, including machine learning, that is designed to approximate a cognitive task; or e. designed to act rationally, including an intelligent software agent or embodied robot that achieves goals using perception, planning, reasoning, learning, communicating, decision making, and acting.Section 1(a) safety testSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a) for artificial intelligence technologyArtificial intelligence technology"Artificial intelligence technology" means a product or service that primarily uses artificial intelligence to perform its intended function.Section 1(a) sold, developed, deployed, used, or offered for sale in this State that is conducted by an artificial intelligence companyArtificial intelligence company"Artificial intelligence company" means a private entity or public agency that sells, develops, deploys, uses, or offers for sale in this State artificial intelligence technology.Section 1(a) pursuant to subsection c. of this section, which requirements shall include but not be limited to: (a) an analysis of potential cybersecurity threats and vulnerabilities; (b) an analysis of an artificial intelligence technologyArtificial intelligence technology"Artificial intelligence technology" means a product or service that primarily uses artificial intelligence to perform its intended function.Section 1(a)'s data sources and potential sources of bias, incorrect or inaccurate information, or violations of State or federal criminal, copyright, patent, or trade secret laws; and (c) descriptions of possible remedies or defensive measures that can be taken by the artificial intelligence companyArtificial intelligence company"Artificial intelligence company" means a private entity or public agency that sells, develops, deploys, uses, or offers for sale in this State artificial intelligence technology.Section 1(a) to address all potential cybersecurity threats and vulnerabilities, potential sources of bias, incorrect or inaccurate information, or potential violations of State or federal criminal, copyright, patent, or trade secret laws identified during the conducting of the safety testSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a);

(2) 1 and (2) review each annual report required to be submitted by an artificial intelligence companyArtificial intelligence company"Artificial intelligence company" means a private entity or public agency that sells, develops, deploys, uses, or offers for sale in this State artificial intelligence technology.Section 1(a) pursuant to subsection c. of this section.

Section 1(b) assigns two duties to the Office of Information Technology. First, the OIT must establish minimum requirements for AI safety tests, which must include at minimum three components: cybersecurity threat and vulnerability analysis, data source and bias analysis (including potential violations of criminal, copyright, patent, or trade secret law), and descriptions of possible remedies or defensive measures. Second, the OIT must review each annual report submitted by an artificial intelligence company under subsection (c).

This section creates obligations on a government agency rather than on regulated private entities, though the standards it produces will shape the compliance burden on AI companies under subsection (c).

Compliance actions 1 item
1
The Office of Information Technology must establish minimum requirements for AI safety testsSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a) — including cybersecurity threat analysis, data source and bias analysis, and remediation descriptions — and must review each annual safety testSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a) report submitted by AI companies.
PS-01.4
Section 1(c)
Annual safety testing and reporting by AI companies
DeployerDeveloper

(c)(1)–(4) 2 An artificial intelligence companyArtificial intelligence company"Artificial intelligence company" means a private entity or public agency that sells, develops, deploys, uses, or offers for sale in this State artificial intelligence technology.Section 1(a) shall annually subject all artificial intelligence technologyArtificial intelligence technology"Artificial intelligence technology" means a product or service that primarily uses artificial intelligence to perform its intended function.Section 1(a) sold, developed, deployed, used, or offered for sale in this State to a safety testSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a) that adheres to the requirements established pursuant to subsection b. of this section and submit a report to the Office of Information Technology containing: (1) a list of all artificial intelligenceArtificial intelligence"Artificial intelligence" means an artificial intelligence system: a. that performs tasks under varying and unpredictable circumstances without significant human oversight or that can learn from experience and improve performance when exposed to data sets; b. developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action; c. designed to think or act like a human, including an artificial intelligence system that has cognitive architectures and neural networks; d. that uses a set of techniques, including machine learning, that is designed to approximate a cognitive task; or e. designed to act rationally, including an intelligent software agent or embodied robot that achieves goals using perception, planning, reasoning, learning, communicating, decision making, and acting.Section 1(a) technologies tested; (2) a description of each safety testSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a) conducted, including the safety testSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a)'s adherence to the requirements established pursuant to subsection b. of this section; (3) a list of all third parties used to conduct safety testsSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a), if any; and (4) the results of each safety testSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a) administered.

Section 1(c) is the bill's central compliance obligation. Every artificial intelligence company must annually conduct safety tests on all AI technology it sells, develops, deploys, uses, or offers for sale in New Jersey, adhering to the minimum requirements the OIT establishes under subsection (b). Each company must then submit a report to OIT containing four enumerated elements: a list of all AI technologies tested, a description of each safety test conducted and its adherence to OIT requirements, a list of any third parties used to conduct the tests, and the results of each test.

The annual cadence and the requirement to test all AI technology in scope make this a potentially sweeping obligation, particularly given the broad definition of artificial intelligence company (which includes users and deployers, not just developers). The bill does not specify penalties for failure to test or report.

Compliance actions 2 items
2
Artificial intelligenceArtificial intelligence"Artificial intelligence" means an artificial intelligence system: a. that performs tasks under varying and unpredictable circumstances without significant human oversight or that can learn from experience and improve performance when exposed to data sets; b. developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action; c. designed to think or act like a human, including an artificial intelligence system that has cognitive architectures and neural networks; d. that uses a set of techniques, including machine learning, that is designed to approximate a cognitive task; or e. designed to act rationally, including an intelligent software agent or embodied robot that achieves goals using perception, planning, reasoning, learning, communicating, decision making, and acting.Section 1(a) companies must annually conduct safety testsSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a) — covering cybersecurity threats, bias, inaccuracies, and legal compliance — on all AI technology they sell, develop, deploy, use, or offer for sale in New Jersey, adhering to OIT-established minimum requirements.
S-01.1
3
Artificial intelligenceArtificial intelligence"Artificial intelligence" means an artificial intelligence system: a. that performs tasks under varying and unpredictable circumstances without significant human oversight or that can learn from experience and improve performance when exposed to data sets; b. developed in computer software, physical hardware, or other context that solves tasks requiring human-like perception, cognition, planning, learning, communication, or physical action; c. designed to think or act like a human, including an artificial intelligence system that has cognitive architectures and neural networks; d. that uses a set of techniques, including machine learning, that is designed to approximate a cognitive task; or e. designed to act rationally, including an intelligent software agent or embodied robot that achieves goals using perception, planning, reasoning, learning, communicating, decision making, and acting.Section 1(a) companies must annually submit to the Office of Information Technology a report listing all AI technologies tested, describing each safety testSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a) conducted and its adherence to OIT requirements, identifying any third parties used for testing, and providing the results of each safety testSafety test"Safety test" means a protocol for assessing artificial intelligence technology for potential biases, inaccuracies, and cybersecurity threats.Section 1(a).
R-02.1
Section 2
Rulemaking authority

The Office of Information Technology shall, in consultation with the Chief Innovation Officer and the Director of the New Jersey Office of Homeland Security and Preparedness, adopt, pursuant to the "Administrative Procedure Act," P.L.1968, c.410 (C.52:14B-1 et seq.), rules and regulations as may be necessary to implement the provisions of P.L. , c. (C. ) (pending before the Legislature as this bill).

Section 2 directs the Office of Information Technology to adopt implementing rules and regulations pursuant to the Administrative Procedure Act, in consultation with the Chief Innovation Officer and the Director of the New Jersey Office of Homeland Security and Preparedness. This is a standard rulemaking delegation that does not itself impose compliance duties on regulated entities.

Section 3
Effective date

This act shall take effect on the first day of the sixth month next following enactment.

Section 3 provides that the act takes effect on the first day of the sixth month following enactment — a standard delayed-effective-date provision giving OIT time to develop regulations and AI companies time to prepare for compliance.

Passage Likelihood

Medium
Status Introduced
Chamber No passage
Committee No action
Majority party (No data)
Bipartisan Yes
Prior session None

Legislative History

2026-03-05 Introduced in the Senate, Referred to Senate Transportation Committee

Entry Last Reviewed

2026-05-20
AI generated