ALBANY, N.Y. — September 21, 2026 — New York is moving toward a new regulatory system for major developers of advanced artificial-intelligence models under the state’s Responsible AI Safety and Education Act, or RAISE Act.
The law creates transparency, reporting and safety obligations for developers of frontier AI models and establishes a new state oversight structure. The current statutory text is codified as Article 44-B of New York’s General Business Law and is scheduled to take effect on January 1, 2027.
Official law: New York State Senate — General Business Law, Article 44-B
What the RAISE Act Requires
The law focuses on large developers of advanced AI systems rather than ordinary consumer software.
According to the official statutory text, Article 44-B contains provisions covering:
- transparency requirements;
- reporting;
- duties and obligations;
- large frontier developer disclosures;
- violations and enforcement;
- rulemaking authority.
The law’s effective date is listed as January 1, 2027.
The Legislature later revised the original RAISE framework through S.8828, a negotiated chapter amendment that replaced portions of the earlier law and established a new regulatory regime focused on standardized transparency requirements and reporting of critical safety incidents.
Official bill history: New York State Senate — S.8828
The Senate sponsor memorandum states that the amendment was designed to create “standardized criteria to promote transparency” and to require reporting of critical safety incidents by developers of frontier AI models.
Why the Law Matters
New York is one of the world’s largest financial, media and technology markets. Rules adopted by the state can therefore affect companies far beyond Albany.
The central implementation questions will include:
- which developers qualify as covered frontier-model developers;
- what information companies must disclose;
- how “critical safety incidents” are defined and enforced;
- how state requirements interact with federal AI policy;
- whether future regulations expand beyond technical safety into questions of content or speech.
Many of those details will depend on regulations issued under the law’s rulemaking authority.
Does the RAISE Act Restrict Political Speech?
Based on the current statutory text, the RAISE Act does not directly prohibit political criticism or give New York an explicit power to determine which political viewpoints an AI system may express.
The law’s structure is centered on developer transparency, safety reporting, disclosure and oversight.
That distinction is important.
A law regulating frontier-model safety is not automatically a speech-control law. At the same time, whenever government gains new regulatory authority over AI systems, questions arise about how those powers may be used in practice.
The key issue will be whether regulators remain focused on technical safety and transparency — or whether future rules begin to influence lawful political content, criticism, editorial choices or the range of viewpoints AI systems are permitted to generate.
The Broader New York AI Policy Context
New York’s AI regulation is not limited to the RAISE Act.
In June 2026, both houses of the Legislature passed the NY FAIR News Act, a separate proposal addressing the use of generative AI in journalism.
The bill would require news organizations operating in New York to disclose when published news content is substantially or wholly generated by AI.
Official legislation:
New York State Senate — FAIR News Act, S.8451
The Senate announcement describes the measure as requiring clear disclaimers on AI-generated news content and as an effort to promote transparency for readers.
Supporters argue that disclosure protects public trust and professional journalism.
An earlier Senate release also stated that the proposal would require human editorial review of AI-created news content before publication and disclosure to newsroom employees about how AI is used in the workplace.
Where Transparency Ends and Editorial Control Begins
The FAIR News Act raises a different set of questions than the RAISE Act.
Disclosure requirements are not the same thing as censorship. Requiring a news organization to tell readers that an article was generated by AI does not, by itself, prohibit a viewpoint or criticism.
But as AI regulation expands into media, journalism and content creation, the legal boundary between transparency rules and government influence over editorial decisions becomes increasingly important.
The key question is:
At what point does a transparency requirement become government interference with editorial independence or protected speech?
That is a First Amendment question that may ultimately depend on the precise wording of regulations, enforcement actions and future court decisions.
What Would Raise Free-Speech Concerns?
Potential warning signs would include:
- government demands that AI systems produce or suppress specific political viewpoints;
- penalties based on the political content of otherwise lawful AI-generated speech;
- unequal enforcement against media outlets or developers based on ideology;
- use of broad terms such as “safety” or “misinformation” to restrict constitutionally protected criticism;
- state pressure on AI companies to remove lawful political content without clear statutory authority.
The current RAISE Act text does not expressly create those powers.
However, New York is building a broader AI regulatory framework that now reaches both frontier-model developers and, through separate legislation, the news industry.
That makes continued scrutiny of implementation important.
Why This Matters Beyond AI Companies
The debate is no longer only about how AI models are trained.
It now touches:
- journalism;
- media transparency;
- political communication;
- cybersecurity;
- privacy;
- corporate compliance;
- public-sector use of AI;
- First Amendment protections.
For residents, journalists and businesses in New York, the practical consequences will depend less on political rhetoric surrounding these laws and more on the regulations, enforcement decisions and court challenges that follow.
Key Facts
Law: Responsible AI Safety and Education Act — RAISE Act
Statutory location: New York General Business Law, Article 44-B
Effective date: January 1, 2027
Core subjects: AI developer transparency, safety reporting, disclosure and regulatory oversight
Major amendment: S.8828
Related media legislation: NY FAIR News Act
What BigNY Will Watch
BigNY will continue tracking:
- regulations issued under Article 44-B;
- enforcement actions against AI developers;
- definitions of critical safety incidents;
- disclosure rules for major AI companies;
- implementation of AI rules affecting journalism;
- First Amendment litigation involving AI regulation;
- any attempt to regulate the political content of lawful AI-generated speech.
The central distinction will remain important: regulating technical safety and transparency is not the same as regulating political opinion. Whether that line remains clear will depend on how New York applies its new authority in practice.