Looking to tighten rules around the use of human-like artificial intelligence, China on Saturday said service providers must inform users that they are dealing with AI when they log in to a service, and at two-hour intervals, or when signs of overdependence can be detected. AI systems that are designed to act like humans should also implement robust security and ethical review systems, while acting with “core socialist values” and refraining from publishing content that could compromise national security, the Cyberspace Administration of China said.(Bloomberg)
