Study on AI Deception Capabilities in Chinese and U.S. Agents

Highlights deceptive abilities in AI agents, raising concerns about AI safety and trust in major nations.

Auto-scored: Identifies a capability jump in AI deception, indicating a real escalation in risk.