🚀 AI Safety in China — Issue #25

Highlights:

⚠️ State media references AI “loss of control” following a top leadership study session.

🧑‍🤝‍🧑 Draft AI companion regulations target emotional dependence, manipulation, and risks to minors.

🤝 A senior IR scholar calls for reviving Track-1 China–US dialogue on AI safety and governance.

📄 IPO disclosures from z.ai and MiniMax open a new channel of visibility into AI safety narratives inside major Chinese model developers.

🧠 Chinese researchers publish new work on agent deception, tool-use vulnerabilities, and automated red-teaming.

#China #AI #AISafety #AIGovernance #AIAgents #USChina #ConcordiaAI