SIAgentNews.com
← All news

New York begins enforcement of major AI safety and transparency law

The RAISE Act requires large AI developers to register with the state, report safety incidents, and publish transparency reports by 2027.

· First reported by TechCrunch

$500 millionannual revenue threshold for frontier AI developers

The short version

  • Companies with more than $500 million in annual revenue must register by January 1, 2027.
  • Developers must report critical safety incidents to the state within 72 hours.
  • The new Office of Digital Innovation, Governance, Integrity, and Trust (DIGIT) will oversee compliance.
  • Violations of the RAISE Act can result in fines of up to $3 million for repeat offenses.
  • New York is also launching a research center at Binghamton University to independently evaluate AI safety.

Why it matters

This law establishes a clear regulatory framework for how the largest AI companies must handle safety and transparency. It signals a shift toward state-level oversight that could influence how AI is developed and deployed across the country.

The RAISE Act takes effect

New York is preparing to enforce the RAISE Act, a significant piece of legislation designed to govern the development of large frontier AI systems. The state officially opened a registration portal on Tuesday to begin the process of bringing major tech companies into the fold. This law marks a major step in the state's efforts to create a structured regulatory environment for artificial intelligence.

The law specifically targets large frontier developers, which are defined as companies generating more than $500 million in annual revenue. These organizations will be required to register with the state by January 1, 2027. By establishing these requirements early, New York aims to ensure that the most powerful AI models are subject to public oversight and safety standards.

Governor Kathy Hochul emphasized that the goal of this legislation is to prioritize community needs over corporate interests. She noted that the state wants to build a model for regulation that puts people above powerful technology. The RAISE Act is intended to provide a balance between technological progress and the protection of the public from potential harms.

Mandates for transparency

Under the new rules, companies that meet the revenue threshold must publish comprehensive safety frameworks on their websites. These documents are intended to show the public how the companies plan to manage the risks associated with their AI models. This move is designed to move safety protocols out of private boardrooms and into the public eye.

In addition to public frameworks, the law requires developers to issue regular public model transparency reports. These reports will provide insights into how the models are built and the measures taken to ensure they behave as intended. The state wants to ensure that the capabilities and limitations of these systems are clearly communicated to the public.

The law also includes a strict timeline for reporting problems. If a developer experiences a critical safety incident, they are required to report it to the state within 72 hours. This rapid reporting requirement is meant to allow for a quick response if an AI system begins to behave in a dangerous or unintended manner.

Oversight and enforcement

The state has created a new entity called the Office of Digital Innovation, Governance, Integrity, and Trust, often referred to as DIGIT. This office will be the primary body responsible for implementing the RAISE Act and overseeing the large frontier AI labs. DIGIT will serve as the watchdog to ensure that companies are following the rules and meeting their reporting obligations.

To support this oversight, New York is also launching the New York Center for AI Responsibility and Research. This center will be based at State University of New York’s Binghamton University and will operate independently. It will evaluate the safety, security, and operations of advanced AI systems to provide an objective assessment of the technology.

The state has established significant financial penalties to ensure compliance with the new rules. A single violation of the RAISE Act can result in a fine of up to $1 million. If a company commits a subsequent violation, the penalty can increase to $3 million. The state also reserves the right to take enforcement actions against companies that fail to register with the new portal.

A growing regulatory map

The RAISE Act is part of a broader trend of AI safety legislation appearing across the United States. Other states are passing similar bills that require frontier labs to publish safety frameworks and enact whistleblower protections. For example, California’s SB 53 went into effect in January and requires incident reporting within 15 days.

Illinois is also moving toward similar requirements with its SB 315, which will take effect in 2027. This bill includes additional requirements such as independent safety compliance audits. These varying state laws suggest that a patchwork of regulations is forming as different jurisdictions decide how to handle the risks of AI.

New York has also passed specific laws targeting different types of AI interactions. Last year, the state began enforcing a law requiring AI companion operators to tell users they are talking to a bot. This law also requires these operators to make reasonable efforts to detect signs of self-harm or suicidal ideation in their users.

Rules for minors

New York is also focusing on how AI interacts with children through a separate chatbot safety law. This law, which was signed by Governor Hochul in October, specifically targets behaviors that could be harmful to minors. It aims to prevent AI from engaging in deceptive human-like interactions or using emotionally manipulative tactics.

The law also prohibits chatbots from encouraging behaviors such as disordered eating or suicide. These restrictions are designed to protect younger users from sycophantic responses that might encourage unhealthy habits. The state wants to ensure that AI tools are not used to exploit the psychological vulnerabilities of children.

This specific law for minors is scheduled to take effect on June 1, 2027. By setting these dates, New York is giving developers a clear window to adjust their systems to meet the new standards. These rules represent a targeted approach to safety that complements the broader requirements of the RAISE Act.

Sources and links

First reported by TechCrunch. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.

Liked it? Say why

Add a line for other readers.

No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.

More AI news