New York's RAISE Act registration portal is now open — and it is one of the strongest AI safety laws in the United States.

The RAISE Act (Responsible AI Safety and Education Act) officially took effect on 10 November 2026. From now on, the largest AI developers must register with the state, publish safety plans, and report major incidents — or face fines of up to $3 million.

Who Must Register

The law targets large frontier AI developers. A company must register if it:

  • Has more than $500 million in annual revenue
  • Has deployed an advanced AI model in New York

Two companies clearly meet that bar: OpenAI and Anthropic. Smaller developers and academic labs are not covered.

Registration Deadline: 1 January 2027

Large developers must register by 1 January 2027 — less than two months after the portal opened.

Registration is only the first step. Companies must also:

  • Publish a safety and security framework — a public document explaining how they test and manage risks before releasing a model.
  • Report critical safety incidents within 72 hours — for example, if a model helps create a bioweapon, carries out a cyberattack on critical infrastructure, or acts outside human control.
  • Publish transparency reports — regular public reports on safety testing, incidents, and model capabilities.

Fines Up to $3 Million

Enforcement is real. The penalties are:

  • Up to $1 million for a first violation
  • Up to $3 million for later violations

The law is enforced by a new state office: the Office of Digital Innovation, Governance, Integrity, and Trust (DIGIT).

What the Companies Said

OpenAI criticised the law in June 2026, calling it "overbroad" and warning it could slow AI development. The company said the safety reporting rules were "vague" and hard to comply with.

Anthropic took a softer line. The company backed parts of the bill — especially transparency and incident reporting — but still raised concerns about how the rules would be enforced.

Both companies have not said yet whether they will register before the January deadline.

Why It Matters

New York is the largest US state to regulate frontier AI so directly. The law could become a template for other states — and a test of whether AI companies will accept binding safety rules.

If the big labs register and publish their safety frameworks, the public will finally get a clearer picture of how frontier models are tested. If they refuse, the fines — and the court fights — will set a precedent for AI regulation across America.

Source: TechCrunch, 10 November 2026.

Originally published on Ai Game Changer Pro.