By mirroring California’s approach, New York reinforces a disclosure-driven model that could become the de facto standard for regulating the most powerful artificial intelligence systems.