TechnologyDeepSeek and other Chinese labs may have used Claude responses to train rival AI
GT Protocol’s digest relays Anthropic’s claim of possible distillation: using Claude-generated responses to train competing systems.
TLDR
GT Protocol says Anthropic claims DeepSeek and other Chinese labs may have used Claude-generated responses to train competing systems through distillation. The digest argues that, if accurate, this raises questions about model sovereignty, intellectual property enforcement and cross-border AI competition.
