New York's landmark artificial intelligence regulation law, the RAISE Act, hasn't taken effect yet, but Governor Kathy Hochul says the state is already looking at ways to bulk it up, including exploring the possibility of "AI kill switches," which would give AI companies or the government the ability to turn off an AI model in an emergency.
On Monday, Governor Hochul held a press conference to announce the state's next steps for implementing the RAISE Act, which goes into effect January 1. The law requires large AI companies working in New York to register with the state, report on their safety plans, and raise the alarm about critical safety incidents within 72 hours.
Hochul also announced the appointment of Marc Gilman, a tech attorney and Fordham professor, as a deputy director in the state's new Office of Digital Innovation, Governance, Integrity, and Trust, or DIGIT, which has the power to propose laws and regulations to improve AI safety. "This technology is changing at warp speed, and it's incumbent upon us to try to get ahead as much as we can," she told reporters.
Monday's announcement was a significant about-face for the governor, who just a year ago succeeded in watering down the RAISE Act from being the nation's leading AI regulatory framework to something akin to what California had passed. For example, the original New York law banned companies from releasing unsafe models—now they only need to provide a warning in their safety plans for the models.