Washington Is Looking to Keep China From Training Its AI on US Models

Anthropic and OpenAI’s warnings about “adversarial distillation” are reinvigorating one of the oldest debates in Silicon Valley.