Tsinghua and Alibaba Experts Discuss Frontier AI Safety and US-China Cooperation
XiaoQianCISS · x · 2026-07-04
This podcast episode features Professor Xu Wei from Tsinghua University and Fu Hongyu from Alibaba Research Institute, delving into core issues surrounding frontier AI safety:
- Safety Thresholds & Evaluations: Discussing whether AI has crossed new capability thresholds in cyber operations and the necessity of mandatory pre-release safety evaluations.
- Open-Weight Risks: Evaluating if open-weight models can remain secure while maintaining openness.
- National Security vs. Public Good: Assessing whether frontier-model safety should be treated as a national security asset or a global public good.
- US-China Cooperation: Identifying practical areas where the US and China can begin cooperating on frontier AI security.