Technology

SB 5 (AN ACT CONCERNING ONLINE SAFETY) requires subscription-based AI providers (e.g., companies offering AI tools via paid plans) to give consumers clear, written disclosures about subscription terms before signing or renewing. This includes detailing any usage limits, such as restrictions based on user behavior or changes to prior terms. The bill also establishes new safety rules for "frontier developers" of advanced AI systems ("foundation models"), defining "catastrophic risk" as scenarios where AI could cause mass harm (e.g., aiding weapon creation or severe physical injury) and mandating risk assessments by covered employees. It does not ban specific AI uses but sets transparency and safety protocols for high-risk systems. The law takes effect October 1, 2026.