New York moved closer Tuesday to becoming the first state to hold developers liable for potential public safety threats from cutting-edge artificial intelligence after lawmakers watered down key legislation with just days left in the session.
AI companies with advanced, large-scale models must create safety plans and disclose safety incidents involving their systems under the legislation (A.B. 6453), which would apply to any developer doing business in the Empire State. Lawmakers amended it late Monday to remove provisions that would have required third-party audits and provided enhanced whistleblower protections.
The Senate will pass the bill on Thursday, the ...