Requires certain artificial intelligence developers to make certain disclosures to Attorney General.

S4446

SENATE, No. 4446

STATE OF NEW JERSEY

222nd LEGISLATURE

�

INTRODUCED JUNE 11, 2026

Sponsored by:

Senator� RAJ MUKHERJI

District 32 (Hudson)

SYNOPSIS

���� Requires certain artificial intelligence developers
to make certain disclosures to Attorney General.

CURRENT VERSION OF TEXT

���� As introduced.

��

An Act
concerning artificial intelligence safety.

����
Be It
Enacted
by the Senate and General Assembly of
the State of New Jersey:

���� 1.��� As used in this act:�

���� �Artificial intelligence technology�
means any technology involving the development or implementation of
machine-based systems that, for explicit or implicit objectives, infer, from
the input it receives, how to generate outputs such as predictions, content,
recommendations, or decisions that may influence physical or virtual
environments.

���� �Catastrophic harm� means the
death or serious injury of 25 or more people or $1 billion in damage to, or
loss of, property arising from a single incident involving a frontier model:�

���� a.���� providing expert-level
assistance in the creation or use of an illegal firearm, lethal autonomous
weapon, chemical weapon, biological weapon, radiological weapon, nuclear
weapon, or explosive device;

���� b.��� engaging in conduct with
no meaningful human oversight, intervention, or supervision that, if it had
been committed by a human, would constitute the crime of murder under
N.J.S.2C:11-3, assault under N.J.S.2C:12-1, or theft under chapter 20 of Title
2C of the New Jersey Statutes, including theft by deception under N.J.S.2C:20-4
and theft by extortion under N.J.S.2C:20-5; or

���� c.���� evading the control of
its frontier developer or user, such as using deceptive techniques against the
frontier developer or user to subvert the controls or monitoring of its
frontier developer or user outside of the context of an evaluation designed to
elicit this behavior.� �Catastrophic harm� shall not include:� incidents
arising from information that a frontier model outputs if the information is
otherwise publicly accessible in a substantially similar form from a source
other than a foundation model; or lawful activity of the federal government.

���� �Catastrophic risk� means a
foreseeable and material risk of catastrophic harm.

���� �Critical safety incident�
means:�

���� a.���� unauthorized access to,
modification of, or exfiltration of, the model weights of a frontier model that
results in death or bodily injury;

���� b.��� harm resulting from the
materialization of a catastrophic risk;

���� c.���� loss of control of a
frontier model causing death or bodily injury; or

���� d.��� a frontier model that
uses deceptive techniques against the frontier developer to subvert the
controls or monitoring of its frontier developer outside of the context of an
evaluation designed to elicit this behavior and in a manner that demonstrates
materially increased catastrophic risk.

���� �Deploying� or �deployment�
means to make a frontier model available to a third party for use,
modification, copying, or combination with other software.

���� �Foundation model� means an
artificial intelligence technology that is trained on a broad data set,
designed for generality of output, and adaptable to a wide range of distinctive
tasks.

���� �Frontier developer� means a
person who has used, or initiated the use of, a quantity of computing power of
at least 10^26 integer or floating-point operations to train a frontier model,
including computing used for the original training run and for any subsequent
fine-tuning, reinforcement learning, or other modifications the person applies.

���� �Frontier model� means a
foundation model that was trained using a quantity of computing power greater
than 10^26 integer or floating-point operations, including computing for the
original training run and for any subsequent fine-tuning, reinforcement
learning, or other material modifications the developer applies to a preceding
foundation model.

����
�G
enerative artificial intelligence system� or �GenAI system� means
a technology system that is trained on data and can generate derived synthetic
content, including text, images, video, and audio, which content emulates the
structure and characteristics of the system�s training data.

���� �Large frontier developer�
means a frontier developer that together with its affiliates collectively had
annual gross revenues in excess of $100 million in the preceding calendar year.

���� �Model weight� means a
numerical parameter in a frontier model that is adjusted through training and
that helps determine how inputs are transformed into outputs.

���� �Office� means the New Jersey
Office of Homeland Security and Preparedness.

���� 2.��� A large frontier
developer who has users in the State of New Jersey shall: �

���� a.���� write and implement
technical and organizational protocols to assess and reduce the risk of the
development, storage, use, or deployment of a frontier model materially
contributing to a catastrophic harm;

���� b.��� annually submit to the
Attorney General a �Risk Management Disclosure� containing a written
explanation of technical and organizational protocols the large frontier developer
has implemented to assess and reduce the risk of the development, storage, use,
or deployment of a frontier model materially contributing to a catastrophic
harm;

���� c.���� annually review and, as
appropriate, update the technical and organizational protocols the large
frontier developer has implemented to assess and reduce the risk of the large
frontier developer�s models materially contributing to a catastrophic harm; and

���� d.��� submit to the Attorney
General a �New Model Risk Disclosure� before, or concurrently with, deploying a
new frontier model or a substantially modified version of an existing frontier
model following the effective date of this act, containing:�

���� (1)�� assessments of
catastrophic risks from the frontier model, described in enough detail that a
third party may exactly replicate the assessments;

���� (2)�� the results of the
assessments required pursuant to paragraph (1) of this subsection;

���� (3)�� the extent to which
third-party evaluators were involved in the assessments required pursuant to
paragraph (1) of this subsection; and

���� (4)�� other steps taken to
fulfill the requirements of the rules issued pursuant to section 3 of this act.

���� 3.��� a.� The Attorney
General, in consultation with the office, shall issue rules requiring that the �Risk
Management Disclosure� submitted pursuant to subsection b. of section 2 of this
act contain language or a visual guide which maps the large frontier developer�s
actions to the suggested actions contained in the most recent edition of the
National Institute of Standards and Technology�s AI Risk Management Framework,
or to a relevant use-case profile based on the framework. �For each suggested
action, the large frontier developer shall state whether the suggested action
is relevant to the large frontier developer�s operations, whether or not the large
frontier developer has adopted the suggested action or a modified version of
the suggested action, a description of how the large frontier developer has
implemented the suggested action or a justification for the large frontier
developer�s decision not to implement the suggested action, including the
criteria the developer used to determine compliance with each suggested action,
and which employees or other persons the developer has made responsible for
implementation of the suggested action.

���� b.��� The Attorney General, in
consultation with the office, may substitute the suggested actions in the National
Institute of Standards and Technology�s AI Risk Management Framework with other
national standards, international standards, or industry-consensus best
practices if the Attorney General, in consultation with the office, determines
that: �

���� (1)�� the AI Risk Management
Framework is out of date; and

���� (2)�� that the National
Institute of Standards and Technology is no longer producing updated editions
of the framework.

���� c.���� The Attorney General
shall, in consultation with the office, issue rules requiring developers to
clearly identify which sections of the �Risk Management Disclosure� submitted
pursuant to subsection b. of section 2 this act were written, edited, or
otherwise contributed to by a generative artificial intelligence system.

���� d.��� The Attorney General
shall publish on Department of Law and Public Safety�s Internet website all disclosures
submitted pursuant to section 2 of this act, with redactions as necessary to
protect the large frontier developer�s trade secrets, the large frontier
developer�s cybersecurity, public safety, the privacy of large frontier
developer�s employees, or the national security of the United States or to
comply with any federal or State law.

���� 4.��� a.� A large frontier
developer who fails to submit any disclosures required pursuant to section 2 of
this act, or who makes materially false or misleading statements in the
frontier developer�s �Risk Management Disclosure� or �New Model Risk Disclosure,�
shall be liable for a civil penalty in the amount of $100,000 for each
violation, which penalty shall be collected by the Attorney General in a
summary proceeding before a court of competent jurisdiction pursuant to the
provisions of the �Penalty Enforcement Law of 1999,� P.L.1999, c.274
(C.2A:58-10 et seq.).

���� b.��� A large frontier
developer who makes materially false or misleading statements about
catastrophic risk in the large frontier developer�s �Risk Management Disclosure�
or �New Model Risk Disclosure� shall not be liable for a civil penalty pursuant
to subsection a. of this section if the statements are found to be reasonable
and made in good faith.

���� 5.��� The Attorney General, in
consultation with the office, may:�

���� a.���� audit a large frontier
developer for compliance with the provisions of this act; or

���� b.��� enter into a contract
with a private entity to conduct an audit pursuant to subsection a. of this
section.

���� 6.��� The Attorney General, in
consultation with the office, shall adopt and promulgate, pursuant to the
�Administrative Procedure Act,� P.L.1968, c.410 (C.52:14B-1 et seq.), rules and
regulations as may be necessary for the implementation of this act.

���� 7.��� No later than one year
prior to the expiration of sections 1 through 5 of this act, the Attorney
General, in consultation with the office, shall submit to the Legislature a
list of recommendations for revisions to the definitions of catastrophic harm,
large frontier developer, artificial intelligence technology, foundation model,
and frontier model, and any other provisions of this act the Attorney General,
in consultation with the office, sees fit.

���� 8.��� This act shall take
effect six months after the date of enactment.� The Attorney General, in
consultation with the office, may take any anticipatory administrative action
in advance as shall be necessary for the implementation of this act.� Sections
1 through 6 of this act shall expire five years after the effective date.�
Section 7 of this act shall expire upon the submission of a list of
recommendations to the Legislature required pursuant to section 7 of this act.

STATEMENT

���� This bill requires certain
artificial intelligence developers to make disclosures to the Attorney
General.� Under the bill, large frontier developers, as that term is defined in
the bill, are required to:� (1) write and implement technical and
organizational protocols to assess and reduce the risk of the development,
storage, use, or deployment of a frontier model materially contributing to a
catastrophic harm; (2) annually submit to the Attorney General a �Risk
Management Disclosure� containing a written explanation of technical and
organizational protocols the large frontier developer has implemented to assess
and reduce the risk of the development, storage, use, or deployment of a
frontier model materially contributing to a catastrophic harm; (3) annually
review and, as appropriate, update the technical and organizational protocols the
large frontier developer has implemented to assess and reduce the risk of the
large frontier developer�s models materially contributing to a catastrophic
harm; and (4) submit to the Attorney General a �New Model Risk Disclosure�
before, or concurrently with, deploying a new frontier model or a substantially
modified version of an existing frontier model, containing assessments of
catastrophic risks from the frontier model, the results of the assessments, the
extent to which third-party evaluators were involved, and other steps taken to
fulfill the requirements of rules issued by the Attorney General, in consultation
with the New Jersey Office of Homeland Security and Preparedness (OHS).

���� Additionally, the bill
requires the Attorney General, in consultation with the OHS, to issue rules
requiring that the �Risk Management Disclosure� submitted by a large frontier
developer contain language or a visual guide which maps the developer�s actions
to the suggested actions contained in the most recent edition of the National
Institute of Standards and Technology�s AI Risk Management Framework, or to a
relevant use-case profile based on the framework, and whether and to what
extent a risk management disclosure was written, edited, or otherwise
contributed to using a generative artificial intelligence system.

���� The bill also states that a
large frontier developer who fails to submit any disclosures required under the
bill, or who makes materially false or misleading statements about catastrophic
risk in the large frontier developer�s �Risk Management Disclosure� or �New Model
Risk Disclosure,� shall be liable for a civil penalty in the amount of $100,000
for each violation.

���� Additionally, the bill
authorizes the Attorney General, in consultation with the OHS, to audit a large
frontier developer for compliance with the provisions of the bill, or enter
into a contract with a private entity to conduct an audit.

���� The bill expires five years
after it takes effect, except that the provision requiring the Attorney
General, in consultation with the OHS, to submit a list of recommendations to
the Legislature expires upon the submission of the list.