New York Responsible AI Safety and Education Act (RAISE Act, S6953B/A6453B)
Effective date
Penalty
$1,000,000 per first violation. $3,000,000 per subsequent violation. AG enforcement. No private right of action.
Obligations mapped
4 obligations
Legislative update
The chapter amendment (S8828) signed March 27, 2026 is the final version. It replaced the original S6953-B entirely, reducing penalties from $10M/$30M to $1M/$3M and adding the DFS Office oversight structure. No further amendments pending.
Overview
If you interact with a frontier AI model (the largest and most powerful AI systems), the company that built it may need to publish a safety framework explaining how it prevents catastrophic risks. Where applicable, they may need to file quarterly risk assessments with New York's DFS Office and report dangerous incidents within 72 hours. This law applies to very large AI companies with over $500 million in revenue. It does not directly create consumer rights, but where applicable it calls for transparency about the safety practices behind the AI systems you use.
This is an AI-specific state law.
See if this regulation applies to your company with the free exposure scan.
Who this applies to
This regulation applies to the following roles:
- Developers and vendors of covered AI systems
- Organizations operating in New York
This regulation applies to companies that build, develop, or sell AI tools, models, or systems. If your company creates AI products that other businesses or consumers use, this regulation may apply to you.
S6953
AI categories covered
- General purpose AI
What this requires you to do
4 obligations identified from statutory analysis.
Article 44-B (chapter amendment)
Article 44-B (chapter amendment)
Article 44-B (chapter amendment)
Article 44-B (chapter amendment)
Regulation summaries are simplified for readability and may not capture every nuance of the underlying statute. Verify important details against primary sources linked on this page.
Enforcement and penalties
$1,000,000 per first violation. $3,000,000 per subsequent violation. AG enforcement. No private right of action.
Penalty amounts are based on statutory text and may be subject to adjustment, judicial interpretation, or enforcement discretion.
Legislative history
effective
Takes effect
guidance issued
Commerce Department identifies NY as having onerous AI laws
guidance issued
Hochul State of the State highlights RAISE Act
signed
Governor Hochul signs RAISE Act (would have been pocket vetoed if unsigned by year-end)
guidance issued
Trump signs EO signaling federal challenge to state AI laws
amended
Amended to S6953B. Passed both chambers.
delayed
Governor Hochul delays signing through December 2025. Chapter amendments negotiated to reduce penalties from $10M/$30M to $1M/$3M.
introduced
S6953 introduced by Sen. Gounardes
Related regulations
- In EffectAI-Specific
California Transparency in Frontier AI Act (SB 53)
Requires developers of frontier AI models trained above the statutory compute threshold (10^26 FLOPs) to publish safety frameworks, report critical safety incidents to the Office of Emergency Services, and implement whistleblower protections. Also reaches large frontier developers with annual revenues over $500 million. Replaces the vetoed SB 1047 with a narrower transparency approach. Currently applies to approximately five to eight companies worldwide given the FLOP threshold. Includes a federal deference provision: compliance with comparable federal standards, including the EU AI Act, is accepted where the statute allows.
Effective
- In EffectAI-Specific
New York Personalized Algorithmic Pricing Disclosure (S 3008, 2025)
Requires businesses to disclose when personalized pricing is set by an algorithm using personal data so consumers know their price may differ from what others see. Statutory date July 8, 2025. The AG paused enforcement during NRF v. James (First Amendment challenge). The court dismissed the challenge October 8, 2025. Enforcement began November 10, 2025. Signed May 9, 2025 as part of omnibus budget bill S.3008. Required disclosure: THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA. Companies that collect protected class data may not combine it with algorithmic pricing to create different prices for different people (anti-discrimination provision beyond disclosure). AG-only enforcement with active enforcement posture (consumer alert issued). Exemptions: regulated financial institutions, state insurance entities, certain subscription contract prices.
Effective
- In EffectAI-Specific
New York AI Companion Models Law (A3008, Article 47)
Requires AI companion operators to disclose AI nature, provide reminders every 3 hours of use, and implement protocols to detect suicidal ideation or self-harm and refer users to crisis services. Carries the highest per-day penalties of any US AI law at $15,000 per day. Penalties fund the NY suicide prevention fund. Unlike California SB 243, the New York law does not include a private right of action.
Effective
- UpcomingAI-Specific
New York Synthetic Performers Disclosure (SB 8420A)
Requires conspicuous disclosure in advertisements when AI-generated synthetic performers are used. A synthetic performer is a digitally created asset generated using AI intended to create the impression of a human performer not recognizable as any real person. Exempt: advertisements for expressive works (movies, TV, streaming, games), audio-only ads, and AI language translation.
Effective
- In EffectAI-Specific
New York Deceased Performer Digital Replicas (SB 8391)
Amends the Right of Publicity law to protect digital replicas of deceased performers. Provides civil remedies for unauthorized AI-generated replicas of deceased individuals for commercial purposes.
Effective
- In EffectAI-Specific
New York Digital Replica Contract Protections (S7676B)
Establishes protections for individuals regarding the use of digital replicas in professional contracts. Requires specific consent provisions for digital replica use in employment and performance contracts.
Effective
New York AI regulation guide lists every tracked rule for this jurisdiction with timelines and obligation tallies.
Regulation summaries are simplified for readability and may not capture every nuance of the underlying statute. Verify important details against primary sources linked on this page.