New York Opens AI Registration and Leaves Audits Outside

Registrations opened on 10 November with 72-hour incident reporting and $3 million penalties, but no independent third-party audit duty.

New York began accepting registrations under its RAISE Act on 10 November 2026, ahead of the law taking effect on 1 January 2027. Large frontier developers, defined as those training models above 10^26 FLOP with more than $500 million in revenue, must publish a safety framework, report critical incidents within 72 hours and file periodic disclosures, with penalties capped at $1 million for a first violation and $3 million thereafter. A state-funded research centre with more than eighty researchers is being stood up to test models independently. The commercial read: registration without an audit is a census rather than a control. It produces a reliable list of who exists, not evidence of what they do, and the compliance budget will flow to self-attestation tooling rather than external verification.