China Tried to Copy Claude. It Ended Up Leaking Its Own Secrets. I China Insights
Anthropic says several Chinese AI companies secretly routed user requests to Claude as part of large-scale efforts to copy its capabilities. Some of those requests contained sensitive data linked to Chinese state-owned companies, police systems and PLA-related activity. AnthropicThe same report also describes China-based defense researchers using Claude for military simulations, including a scenario involving 12 targets in Taiwan. We break down how AI “distillation” works—and how China’s race to catch up in AI may be creating a security problem of its own. Anthropic#Anthropic #ClaudeAI…
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.