US data vendor sells training data to Chinese AI labs
What happened
On November 11, The Decoder reported on a ChinaTalk investigation: at least nine US data companies sold training data to Chinese AI labs, in deals worth about $500 million, including reinforcement learning environments. The Chinese labs bought the same kinds of products previously sold to Anthropic or OpenAI, to cut in-house development and wasted training, and combined them with GPT and Claude outputs for knowledge distillation. Such environments are largely outside export controls today. The investigation's author argues that coding and biology environments with dual-use risk should be brought under existing controls.
Written by AI from the coverage · updated 13 hours ago
Coverage
Follow the reports to see the story from different sides.
- The DecoderUS data brokers are selling half a billion dollars in RL training environments to Chinese AI labs
ChinaTalk 调查发现，至少9家美国数据公司向中国 AI 实验室出售训练数据，交易规模约5亿美元，其中包括强化学习环境。中国实验室购买此前售予 Anthropic 或 OpenAI 的同类产品，以减少内部开发和无效训练，并结合 GPT、Claude 输出进行知识蒸馏。此类环境目前基本不受出口管制，调查作者主张将具有军民两用风险的编码、生物学环境纳入现有管制。
Heat over time
Heat now 7·Comparable peak 10（Nov 11 23:00）·Comparable change over 24 hours –
The trend only compares accounts observed without gaps, so its range may be smaller than the current heat. Hover or tap the chart for each hour; the left and right arrow keys step through it.