New York is moving to enforce a sweeping AI safety law that places new obligations on major developers including OpenAI and Anthropic. Under the law, these companies must register with the state, disclose their safety practices, and report critical incidents, with fines on the table for those that fail to comply.

The law represents one of the most aggressive attempts by a state government to regulate frontier AI development. It targets the largest developers directly, requiring transparency about how their systems are built and monitored, and creating a formal channel for reporting serious problems when they arise. The registration requirement means affected companies must identify themselves to state authorities rather than operating without a formal presence in regulatory records, while the disclosure provisions open their internal safety practices to outside scrutiny. The incident reporting mandate goes further, turning what would once have been internal matters into events subject to state review.

The requirements arrive as AI companies face growing scrutiny from lawmakers at both the state and federal level. New York's approach could serve as a model for other states considering similar legislation, and its enforcement will be watched closely by the industry. How aggressively state officials pursue penalties against developers that fall short of the requirements will shape whether other legislatures follow New York's lead or opt for lighter-touch frameworks.

For OpenAI and Anthropic, the law adds a new layer of compliance work and public disclosure. The companies will need to formalize their safety reporting and registration processes, and any critical incidents will now be subject to state review rather than handled entirely internally. That shift carries significance beyond paperwork: incidents that companies might previously have resolved behind closed doors will instead enter a formal state channel, where regulators can examine how they were handled and whether the developers' disclosed safety practices match their conduct in practice.