China open-weight AI safety report proposes 6-stage risk management process

Chinese AI developers, led by Z.ai and Concordia AI, released a report proposing a six-stage risk management process for open-weight AI models, aiming to balance safety with openness. The report is the first comprehensive framework of its kind from China, signaling a policy shift toward proactive governance. The impact is uncertain as it is a proposal, not regulation, but it could shape future Chinese AI export and deployment policies.