Overview
- The law requires large AI developers to publish safety protocols and to report critical incidents within 72 hours, including when they reasonably believe one has occurred.
- New York will establish a state oversight office to monitor AI development and compliance under the transparency regime.
- Fines are capped at up to $1 million for a first violation and $3 million for subsequent violations, reduced from earlier proposals of up to $10 million and $30 million.
- Legislators dropped provisions to use computational cost as a coverage trigger and to prohibit releasing unsafe models, opting instead for warnings within safety plans.
- New York becomes the second major state with an AI safety law after California’s, as President Trump’s executive order instructs federal agencies to challenge such state rules; OpenAI and Anthropic backed the transparency focus while calling for federal legislation.