|
|
| (One intermediate revision by the same user not shown) |
| Line 1: |
Line 1: |
| '''New York RAISE Act Amendments (S 8828)''' — Governor Kathy Hochul signed Senate Bill S 8828 on '''March 27, 2026''', amending the Responsible AI Safety and Education (RAISE) Act to align New York's frontier AI transparency requirements more closely with California's approach, narrowing the definition of covered developers, reducing penalties, and delaying enforcement to January 1, 2027.<ref name="wiley">[https://www.wiley.law/alert-New-York-Finalizes-RAISE-Act-for-Frontier-AI-Models-Law-Takes-Effect-January-1-2027 Wiley, "New York Finalizes RAISE Act for Frontier AI Models; Law Takes Effect January 1, 2027," April 2026]</ref><ref name="mofo">[https://www.mofo.com/resources/insights/260403-new-york-amends-the-raise-act-to-align-more-closely Morrison & Foerster, "New York Amends the RAISE Act to Align More Closely with California's TFAIA," April 3, 2026]</ref><ref name="nelsonmullins">[https://www.nelsonmullins.com/insights/alerts/privacy_and_data_security_alert/all/new-york-laws-raise-the-bar-in-addressing-ai-safety-the-raise-act-and-ai-companion-models Nelson Mullins, "New York Laws Raise the Bar in Addressing AI Safety: The RAISE Act and AI Companion Models," 2026]</ref>
| | #REDIRECT [[News April 2026]] |
| | |
| == Background: Original RAISE Act ==
| |
| | |
| The original RAISE Act (S 6953-b/A 6453-b) was signed by Governor Hochul on '''December 19, 2025''', making New York the second state after California to impose transparency and safety requirements on developers of advanced AI models.<ref name="hochulsigns">[https://www.governor.ny.gov/news/governor-hochul-signs-nation-leading-legislation-require-ai-frameworks-ai-frontier-models Governor Hochul, "Governor Hochul Signs Nation-Leading Legislation to Require AI Frameworks for Frontier Models," December 19, 2025]</ref><ref name="dlapiper">[https://www.dlapiper.com/insights/publications/2025/12/the-raise-act-new-york-joins-california-in-requiring-developer-transparency-for-large-ai-models DLA Piper, "The RAISE Act: New York Joins California in Requiring Developer Transparency for Large AI Models," December 2025]</ref>
| |
| | |
| The original law applied to developers of AI models trained using more than 10<sup>26</sup> computational operations and established transparency, safety, and incident reporting requirements.
| |
| | |
| == S 8828 Amendments ==
| |
| | |
| Senate Bill S 8828, introduced by Senator Andrew Gounardes on January 8, 2026, made significant amendments to the RAISE Act:<ref name="nysenate">[https://www.nysenate.gov/legislation/bills/2025/S8828 New York Senate, S8828 Bill Page]</ref><ref name="wiley" /><ref name="mofo" />
| |
| | |
| === Narrowed Developer Definition ===
| |
| The amended law narrows the definition of "large frontier developer" from a compute-based threshold to a '''$500 million annual gross revenue threshold''' (including affiliates) — matching California's approach under SB 53 (TFAIA).<ref name="mofo" /> This significantly limits the number of AI companies subject to the law's requirements.
| |
| | |
| === Reduced Penalties ===
| |
| The amendments reduce civil penalties from the original $10 million (first violation) and $30 million (subsequent violations) to '''$1 million''' (first violation) and '''$3 million''' (subsequent violations). Non-filing of disclosures incurs a penalty of $1,000 per day.<ref name="nelsonmullins" /><ref name="mofo" />
| |
| | |
| === Delayed Enforcement ===
| |
| Enforcement was pushed back to '''January 1, 2027''', giving covered developers additional time to comply.<ref name="wiley" /><ref name="mofo" />
| |
| | |
| === Key Provisions Preserved ===
| |
| The core transparency and safety requirements remain intact:<ref name="nelsonmullins" /><ref name="ktslaw">[https://ktslaw.com/en/insights/alert/2026/1/new%20yorks-raise-act-raises-the-bar-for-frontier-ai-developers KTS Law, "New York's RAISE Act Raises the Bar for Frontier AI Developers," January 2026]</ref>
| |
| | |
| * '''Oversight Office''': Establishes an Office within the New York Department of Financial Services for oversight of frontier AI developers, including two reporting channels for critical safety incidents
| |
| * '''Disclosure Requirements''': Large frontier developers must file disclosures including business identity, principal place of business, New York offices, and (for private/closely held companies) lists of entities owning ≥5% interest
| |
| * '''Safety and Security Protocols''': Developers must make safety and security protocols available to authorities
| |
| * '''Incident Reporting''': Developers must report critical safety incidents through the oversight office
| |
| * '''Academic Exemptions''': Accredited colleges, universities, and the Empire AI Consortium and Institute are exempt
| |
| * '''Whistleblower Protections''': Employee protections for reporting safety concerns
| |
| | |
| === Alignment with California ===
| |
| The amendments bring New York's RAISE Act closer to California's Transparency in Frontier AI Act (SB 53 / TFAIA), including the revenue-based threshold and reduced penalties. Key differences remain: New York includes academic exemptions that California lacks, and New York's scope is limited to activities within the state while California may apply extraterritorially.<ref name="mofo" />
| |
| | |
| == Legislative Timeline ==
| |
| | |
| * '''January 8, 2026''': S 8828 introduced by Senator Gounardes<ref name="nysenate" />
| |
| * '''January 28, 2026''': Passed Senate (58-1)<ref name="nysenate" />
| |
| * '''March 11, 2026''': Passed Assembly<ref name="nysenate" />
| |
| * '''March 20, 2026''': Delivered to Governor<ref name="nysenate" />
| |
| * '''March 27, 2026''': Signed into law as Chapter 96<ref name="nysenate" /><ref name="wiley" />
| |
| | |
| == Significance ==
| |
| | |
| New York's amended RAISE Act, together with California's TFAIA, establishes a two-state framework for frontier AI transparency that covers the nation's two largest AI industry hubs. The shift from a compute-based to a revenue-based threshold means the law applies to fewer but more financially significant AI developers, reducing compliance burdens on smaller organizations while maintaining oversight of the largest frontier model operators.
| |
| | |
| The law represents a significant state-level counterweight to federal preemption efforts, particularly the Trump administration's National Policy Framework for AI, which advocates for overriding state AI regulations.<ref name="mofo" />
| |
| | |
| == See Also ==
| |
| | |
| * [[News California-SB-1000-AI-Content-Provenance-2026|California SB 1000 (AI Content Provenance)]]
| |
| * [[News White-House-National-Policy-Framework-AI-2026|White House National Policy Framework for AI]]
| |
| * [[Legislation]]
| |
| | |
| == References ==
| |
| | |
| <references />
| |
| | |
| [[Category:State Legislation]]
| |
| [[Category:New York]]
| |
| [[Category:Federal Preemption]]
| |
| [[Category:Transparency]]
| |
| [[Category:Risk-Based Framework]]
| |
