New York begins regulatory registration for large AI model builders

New York state has launched an official registration system for major artificial intelligence developers as part of its upcoming safety legislation known as the RAISE Act. According to reporting by TechCrunch, tech firms that generate over 500 million dollars in annual revenue, such as OpenAI and Anthropic, must register before 2027. Under the new rules, these businesses must publish safety policies on their websites and inform state regulators about severe safety incidents within 72 hours.

The initiative will be supervised by a new state office named DIGIT, with penalties reaching up to 3 million dollars for repeated non-compliance. To assist with technical oversight, Governor Kathy Hochul established a 55 million dollar research facility at Binghamton University, which will evaluate advanced models and examine corporate safety standards.

For everyday consumers, this development means state authorities are building concrete guardrails around high-powered software that powers consumer tools. In the absence of comprehensive federal laws, individual states such as California, Illinois, and New York are establishing their own accountability rules. This regional approach forces major tech developers to comply with public safety standards if they want to do business in key markets.

However, it remains unclear how state authorities will successfully detect whether a company fails to disclose a safety flaw. Relying on self-reporting within a tight 72-hour window may prove difficult to enforce without strong whistleblower incentives or continuous technical auditing.