STATE OF NEW YORK
________________________________________________________________________
6953--A
2025-2026 Regular Sessions
IN SENATE
March 27, 2025
___________
Introduced by Sens. GOUNARDES, BRISPORT, FAHY, HARCKHAM, JACKSON, KRUEG-
ER, LIU, MAYER, PALUMBO, SALAZAR -- read twice and ordered printed,
and when printed to be committed to the Committee on Internet and
Technology -- committee discharged, bill amended, ordered reprinted as
amended and recommitted to said committee
AN ACT to amend the general business law, in relation to the training
and use of artificial intelligence frontier models
The People of the State of New York, represented in Senate and Assem-bly, do enact as follows:
1 Section 1. Short title. This act shall be known and may be cited as
2 the "Responsible AI safety and education act" or "RAISE act".
3 § 2. The general business law is amended by adding a new article 44-B
4 to read as follows:
5 ARTICLE 44-B
6 RESPONSIBLE AI SAFETY AND EDUCATION (RAISE) ACT
7 Section 1420. Definitions.
8 1421. Transparency requirements regarding frontier model train-
9 ing and use.
10 1422. Protections, rights and obligations of employees.
11 1423. Violations.
12 1424. Duties and obligations.
13 1425. Scope.
14 1426. Severability.
15 § 1420. Definitions. As used in this article, the following terms
16 shall have the following meanings:
17 1. "Appropriate redactions" means redactions to a safety and security
18 protocol or audit report that a developer may make when necessary to:
19 (a) protect public safety to the extent the developer can reasonably
20 predict such risks;
21 (b) protect trade secrets;
EXPLANATION--Matter in italics (underscored) is new; matter in brackets
[] is old law to be omitted.
LBD00047-13-5
S. 6953--A 2
1 (c) prevent the release of confidential information as required by
2 state or federal law;
3 (d) protect employee or customer privacy; or
4 (e) prevent the release of information otherwise controlled by state
5 or federal law.
6 2. "Artificial intelligence" means a machine-based system that can,
7 for a given set of human-defined objectives, make predictions, recommen-
8 dations, or decisions influencing real or virtual environments, and that
9 uses machine- and human-based inputs to perceive real and virtual envi-
10 ronments, abstract such perceptions into models through analysis in an
11 automated manner, and use model inference to formulate options for
12 information or action.
13 3. "Artificial intelligence model" means an information system or
14 component of an information system that implements artificial intelli-
15 gence technology and uses computational, statistical, or machine-learn-
16 ing techniques to produce outputs from a given set of inputs.
17 4. "Compute cost" means the cost incurred to pay for compute used in
18 training a model when calculated using the average published market
19 prices of cloud compute in the United States at the start of training
20 such model as reasonably assessed by the person doing the training.
21 5. "Deploy" means to use a frontier model or to make a frontier model
22 foreseeably available to one or more third parties for use, modifica-
23 tion, copying, or a combination thereof with other software, except for
24 training or developing the frontier model, evaluating the frontier model
25 or other frontier models, or complying with federal or state laws.
26 6. "Frontier model" means either of the following:
27 (a) an artificial intelligence model trained using greater than 10§26
28 computational operations (e.g., integer or floating-point operations),
29 the compute cost of which exceeds one hundred million dollars; or
30 (b) an artificial intelligence model produced by applying knowledge
31 distillation to a frontier model as defined in paragraph (a) of this
32 subdivision.
33 7. "Critical harm" means the death or serious injury of one hundred or
34 more people or at least one billion dollars of damages to rights in
35 money or property caused or materially enabled by a large developer's
36 creation, use, storage, or release of a frontier model, through either
37 of the following:
38 (a) The creation or use of a chemical, biological, radiological, or
39 nuclear weapon; or
40 (b) An artificial intelligence model engaging in conduct that does
41 both of the following:
42 (i) Acts with limited human intervention; and
43 (ii) Would, if committed by a human, constitute a crime specified in
44 the penal law that requires intent, recklessness, or gross negligence,
45 or the solicitation or aiding and abetting of such a crime.
46 A harm inflicted by an intervening human actor shall not be deemed to
47 result from a developer's activities unless such activities made it
48 substantially easier or more likely for the actor to inflict such harm.
49 8. "Knowledge distillation" means any supervised learning technique
50 that uses a larger artificial intelligence model or the output of a
51 larger artificial intelligence model to train a smaller artificial
52 intelligence model with similar or equivalent capabilities as the larger
53 artificial intelligence model.
54 9. "Large developer" means a person that has trained at least one
55 frontier model, the compute cost of which exceeds five million dollars,
56 and has spent over one hundred million dollars in compute costs in
S. 6953--A 3
1 aggregate in training frontier models. Accredited colleges and univer-
2 sities shall not be considered large developers under this article to
3 the extent that such colleges and universities are engaging in academic
4 research. If a person subsequently transfers full intellectual property
5 rights of the frontier model to another person (including the right to
6 resell the model) and retains none of those rights for themself, then
7 the receiving person shall be considered the large developer and shall
8 be subject to the responsibilities and requirements of this article
9 after such transfer.
10 10. "Model weight" means a numerical parameter in an artificial intel-
11 ligence model that is adjusted through training and that helps determine
12 how inputs are transformed into outputs.
13 11. "Person" means an individual, proprietorship, firm, partnership,
14 joint venture, syndicate, business trust, company, corporation, limited
15 liability company, association, committee, or any other nongovernmental
16 organization or group of persons acting in concert.
17 12. "Safety and security protocol" means documented technical and
18 organizational protocols that:
19 (a) Specify reasonable protections and procedures that, if successful-
20 ly implemented would appropriately reduce the risk of critical harm;
21 (b) Describe reasonable administrative, technical, and physical
22 cybersecurity protections for frontier models within the large develop-
23 er's control that, if successfully implemented, appropriately reduce the
24 risk of unauthorized access to, or misuse of, the frontier models lead-
25 ing to critical harm, including by sophisticated actors;
26 (c) Describe in detail the testing procedure to evaluate if the fron-
27 tier model poses an unreasonable risk of critical harm;
28 (d) Describe in detail how the testing procedure assesses whether the
29 frontier model could be misused, be modified, be executed with increased
30 computational resources, evade the control of its large developer or
31 user, be combined with other software or be used to create another fron-
32 tier model in a manner that would increase the risk of critical harm;
33 (e) State compliance requirements with sufficient detail and specif-
34 icity to allow the large developer or a third party to readily ascertain
35 whether the requirements of the safety and security protocol have been
36 followed;
37 (f) Describe how the large developer will fulfill their obligations
38 under this article, including with respect to any requirements, safe-
39 guards, or modifications; and
40 (g) Designate senior personnel to be responsible for ensuring compli-
41 ance.
42 13. "Safety incident" means an incident of the following kinds that
43 occurs in such a way that it provides demonstrable evidence of an
44 increased risk of critical harm:
45 (a) A frontier model autonomously engaging in behavior other than at
46 the request of a user;
47 (b) Theft, misappropriation, malicious use, inadvertent release, unau-
48 thorized access, or escape of the model weights of a frontier model;
49 (c) The critical failure of any technical or administrative controls,
50 including controls limiting the ability to modify a frontier model; or
51 (d) Unauthorized use of a frontier model.
52 14. "Trade secret" means any form and type of financial, business,
53 scientific, technical, economic, or engineering information, including a
54 pattern, plan, compilation, program device, formula, design, prototype,
55 method, technique, process, procedure, program, or code, whether tangi-
56 ble or intangible, and whether or how stored, compiled, or memorialized
S. 6953--A 4
1 physically, electronically, graphically, photographically or in writing,
2 that:
3 (a) Derives independent economic value, actual or potential, from not
4 being generally known to, and not being readily ascertainable by proper
5 means by, other persons who can obtain economic value from its disclo-
6 sure or use; and
7 (b) Is the subject of efforts that are reasonable under the circum-
8 stances to maintain its secrecy.
9 § 1421. Transparency requirements regarding frontier model training
10 and use. 1. Before deploying a frontier model, the large developer of
11 such frontier model shall do all of the following:
12 (a) Implement a written safety and security protocol;
13 (b) Retain an unredacted copy of the safety and security protocol,
14 including records and dates of any updates or revisions. Such unredacted
15 copy of the safety and security protocol, including records and dates of
16 any updates or revisions, shall be retained for as long as a frontier
17 model is deployed plus five years;
18 (c) (i) Conspicuously publish a copy of the safety and security proto-
19 col with appropriate redactions and transmit a copy of such redacted
20 safety and security protocol to the division of homeland security and
21 emergency services;
22 (ii) Grant the division of homeland security and emergency services or
23 the attorney general access to the safety and security protocol, with
24 redactions only to the extent required by federal law, upon request;
25 (d) Record, as and when reasonably possible, and retain for as long as
26 the frontier model is deployed plus five years information on the
27 specific tests and test results used in any assessment of the frontier
28 model that provides sufficient detail for third parties to replicate the
29 testing procedure; and
30 (e) Implement appropriate safeguards to prevent unreasonable risk of
31 critical harm.
32 2. A large developer shall not deploy a frontier model if doing so
33 would create an unreasonable risk of critical harm.
34 3. A large developer shall conduct an annual review of any safety and
35 security protocol required by this section to account for any changes
36 to the capabilities of their frontier models and industry best practices
37 and, if necessary, make modifications to such safety and security proto-
38 col. If any modifications are made, the large developer shall publish
39 the safety and security protocol in the same manner as required pursuant
40 to paragraph (c) of subdivision one of this section.
41 4. (a) Beginning on the effective date of this article, or ninety days
42 after a developer first qualifies as a large developer, whichever is
43 later, a large developer shall annually retain a third party to perform
44 an independent audit of compliance with the requirements of this
45 section. Such third party shall conduct audits consistent with best
46 practices.
47 (b) The third party shall be granted access to unredacted materials as
48 necessary to comply with the third party's obligations under this subdi-
49 vision.
50 (c) The third party shall produce a report including all of the
51 following:
52 (i) A detailed assessment of the large developer's steps to comply
53 with the requirements of this section;
54 (ii) If applicable, any identified instances of noncompliance with the
55 requirements of this section, and any recommendations for how the devel-
S. 6953--A 5
1 oper can improve its policies and processes for ensuring compliance with
2 the requirements of this section;
3 (iii) A detailed assessment of the large developer's internal
4 controls, including its designation and empowerment of senior personnel
5 responsible for ensuring compliance by the large developer, its employ-
6 ees, and its contractors; and
7 (iv) The signature of the lead auditor certifying the results of the
8 audit.
9 (d) The large developer shall retain an unredacted copy of the report
10 for as long as a frontier model is deployed plus five years.
11 (e) (i) The large developer shall conspicuously publish a copy of the
12 third party's report with appropriate redactions and transmit a copy of
13 such redacted report to the division of homeland security and emergency
14 services.
15 (ii) The large developer shall grant the division of homeland security
16 and emergency services or the attorney general access to the third
17 party's report, with redactions only to the extent required by federal
18 law, upon request.
19 5. A large developer shall disclose each safety incident affecting
20 the frontier model to the division of homeland security and emergency
21 services within seventy-two hours of the large developer learning of the
22 safety incident or within seventy-two hours of the large developer
23 learning facts sufficient to establish a reasonable belief that a safety
24 incident has occurred. Such disclosure shall include: (a) the date of
25 the safety incident; (b) the reasons the incident qualifies as a safety
26 incident as defined in subdivision thirteen of section fourteen hundred
27 twenty of this article; and (c) a short and plain statement describing
28 the safety incident.
29 6. A large developer shall not knowingly make false or materially
30 misleading statements or omissions in or regarding documents produced
31 pursuant to this section.
32 7. Any person who is not a large developer, but who sets out to train
33 a frontier model that if completed as planned would qualify such person
34 as a large developer (i.e. at the end of the training, such person will
35 have spent five million dollars in compute costs on one frontier model
36 and one hundred million dollars in compute costs in aggregate in train-
37 ing frontier models, excluding accredited colleges and universities to
38 the extent such colleges and universities are engaging in academic
39 research) shall, before training such model:
40 (a) Implement a written safety and security protocol, excluding the
41 requirements described in paragraphs (c) and (d) of subdivision twelve
42 of section fourteen hundred twenty of this article; and
43 (b) Transmit a copy of an appropriately redacted safety and security
44 protocol to the division of homeland security and emergency services.
45 § 1422. Protections, rights and obligations of employees. 1. A large
46 developer or a contractor or subcontractor of a large developer shall
47 not prevent an employee from disclosing, or threatening to disclose, or
48 retaliate against an employee for disclosing or threatening to disclose,
49 information to the large developer or the attorney general, if the
50 employee has reasonable cause to believe that the large developer's
51 activities pose an unreasonable or substantial risk of critical harm,
52 regardless of the employer's compliance with applicable law.
53 2. An employee harmed by a violation of this section may petition a
54 court for appropriate temporary or preliminary injunctive relief.
55 3. A large developer shall inform employees of their protections,
56 rights and obligations under this article within ninety days of the
S. 6953--A 6
1 effective date of this article or of becoming a large developer, which-
2 ever is later, upon commencement of employment, and by posting a notice
3 thereof. Such notice shall be posted conspicuously in easily accessible
4 and well-lighted places customarily frequented by employees.
5 4. Nothing in this section shall be deemed to diminish the rights,
6 privileges, or remedies of any employee under any other law or regu-
7 lation or under any collective bargaining agreement or employment
8 contract.
9 5. As used in this section, the following terms shall have the follow-
10 ing meanings:
11 (a) "Employee" has the same meaning as defined in subdivision five of
12 section two of the labor law and includes both of the following:
13 (i) Contractors or subcontractors and unpaid advisors involved with
14 assessing, managing, or addressing the risk of critical harm from fron-
15 tier models; and
16 (ii) Corporate officers.
17 (b) "Contractor or subcontractor" means any person, sole proprietor,
18 partnership, firm, corporation, limited liability company, association
19 or other legal entity who by oneself or through others offers to under-
20 take, or holds oneself out as being able to undertake, or does undertake
21 work assessing, managing, or addressing the risk of critical harm from
22 frontier models on behalf of the large developer.
23 § 1423. Violations. 1. The attorney general may bring a civil action
24 for a violation of this article and to recover all of the following:
25 (a) For a violation of section fourteen hundred twenty-one of this
26 article, a civil penalty in an amount not exceeding ten million dollars
27 for a first violation and in an amount not exceeding thirty million
28 dollars for any subsequent violation.
29 (b) For a violation of section fourteen hundred twenty-two of this
30 article, a civil penalty in an amount not exceeding ten thousand dollars
31 per employee for each violation of such section to be awarded to the
32 employee who was retaliated against.
33 (c) For a violation of section fourteen hundred twenty-one or fourteen
34 hundred twenty-two of this article, injunctive or declaratory relief.
35 2. (a) A provision within a contract or agreement that seeks to waive,
36 preclude, or burden the enforcement of a liability arising from a
37 violation of this article, or to shift that liability to any person or
38 entity in exchange for their use or access of, or right to use or
39 access, a large developer's products or services, including by means of
40 a contract of adhesion, is void as a matter of public policy.
41 (b) A court shall disregard corporate formalities and impose joint and
42 several liability on affiliated entities for purposes of effectuating
43 the intent of this section to the maximum extent allowed by law if the
44 court concludes that both of the following are true:
45 (i) The affiliated entities, in the development of the corporate
46 structure among the affiliated entities, took steps to purposely and
47 unreasonably limit or avoid liability; and
48 (ii) As the result of the steps described in subparagraph (i) of this
49 paragraph, the corporate structure of the large developer or affiliated
50 entities would frustrate recovery of penalties, damages, or injunctive
51 relief under this section.
52 3. The division of homeland security and emergency services shall
53 make any critical safety incident disclosure available to the attorney
54 general upon request.
55 4. This section does not limit the application of other laws.
S. 6953--A 7
1 § 1424. Duties and obligations. The duties and obligations imposed by
2 this article are cumulative with any other duties or obligations imposed
3 under other law and shall not be construed to relieve any party from any
4 duties or obligations imposed under other law and do not limit any
5 rights or remedies under existing law.
6 § 1425. Scope. This article shall only apply to frontier models that
7 are developed, deployed, or operating in whole or in part in New York
8 state.
9 § 1426. Severability. If any clause, sentence, paragraph, subdivision,
10 section or part of this article shall be adjudged by any court of compe-
11 tent jurisdiction to be invalid, such judgment shall not affect, impair,
12 or invalidate the remainder thereof, but shall be confined in its opera-
13 tion to the clause, sentence, paragraph, subdivision, section, or part
14 thereof directly involved in the controversy in which such judgment
15 shall have been made.
16 § 3. This act shall take effect on the ninetieth day after it shall
17 have become a law.
