New York Enforces AI Safety Law, Targets OpenAI and Anthropic
New York enforces AI safety law requiring major AI developers to register by 2027.
Why it matters: This enforcement sets a precedent for AI regulation, affecting legal compliance and tech firms nationwide. Legal professionals must adapt to new disclosure and safety obligations.
- RAISE Act applies to AI firms with $500M+ revenue and foundation models of 10^26 operations or more.
- Registration deadline set for January 1, 2027; firms must publish AI safety frameworks and report critical incidents within 72 hours.
- DIGIT enforces compliance with fines up to $1M initially and $3M for repeat violations.
- New York invests $55M in Center for AI Responsibility and Research at Binghamton University to support oversight and compliance.
New York has begun enforcing the Responsible AI Safety and Education Act (RAISE Act), a pioneering law regulating large AI developers such as OpenAI and Anthropic. The law applies to companies with over $500 million in annual revenue that develop "foundation models"—core AI systems trained with at least 1026 computational operations, a measure of model complexity and scale.
Under RAISE, qualifying AI firms must register with New York's new oversight office, the Office of Digital Innovation, Governance, Integrity, and Trust (DIGIT), by January 1, 2027. They are also required to publicly disclose AI safety frameworks explaining how their systems minimize risks, and to report any serious safety incidents within 72 hours. This reporting window tightens to 24 hours when there is a risk of death or serious harm.
DIGIT is the agency responsible for enforcement and may impose fines of up to $1 million for initial violations and up to $3 million for repeated offenses. Governor Kathy Hochul emphasized the law’s focus on community safety, stating, "We’re building a national model for smart, sensible AI regulation, a model that puts communities above corporations and people above powerful technology."
New York has allocated $55 million to create the New York Center for AI Responsibility and Research at Binghamton University. This center will independently assess AI systems and assist DIGIT in compliance monitoring and enforcement.
The RAISE Act complements similar legislation such as California’s Transparency in Frontier AI Act, reflecting a broader trend of state-level AI regulation. In parallel, the New York City Council has issued subpoenas to OpenAI and Anthropic, demanding testimony about AI safety risks, signaling rising oversight scrutiny.
These regulatory developments carry important implications for legal and compliance professionals advising AI developers and legal tech firms, who must now navigate evolving rules for AI risk management and transparency reporting.
By the numbers:
- $500 million+ revenue — threshold for AI firms under RAISE Act
- January 1, 2027 — deadline for AI firms to register with New York
- $1 million and $3 million — maximum fines for initial and repeat violations
What's next: Enforcement begins immediately with reporting obligations; firms should prepare for stricter incident disclosure rounds by 2027.