S-10701, called the “TERMINATOR Act” in the bill text, would amend New York’s General Business Law, including provisions added by a 2026 law, and add new safety rules for frontier AI models. Its stated goal is to strengthen safety evaluation, monitoring, reporting, security, employee protections and enforcement; those rules would not guarantee that models are safe. The bill is not law, and its passage is uncertain.
The bill requires developers to pay evaluation costs and be assessed for administration, but the available material gives no dollar estimates.
Bills on related subjects, starting with New York.
Large frontier AI developers would have to publish a safety framework and hire independent examiners. The measure would require incident reporting within 72 hours, with penalties up to $1 million.
Status
REFERRED TO RULES
Senate bill would impose strict liability on AI developers for injuries from 'covered models' meeting high compute and cost thresholds, taking effect immediately if enacted.
Status
REFERRED TO RULES
If passed, lawmakers must reveal when AI helped draft remarks and include a percentage, making official speeches more transparent.
Status
REFERRED TO RULES
Lawmakers must disclose AI involvement before remarks. The record would show a digital warning if disclosure is missing.
Status
referred to governmental operations