白宫科技政策办公室将推动资讯共享，协助业界侦测未经授权的模型提取行为，并与产业合作制定应对措施，遏制此类滥用行为并追究不良行为者的责任。
美国开发商（如 OpenAI、Anthropic 和 Google）对中国竞争对手（如 DeepSeek）利用大规模帐户进行「工业级」蒸馏、规避安全限制并降低开发成本的行为表示严重关切。
The Trump administration has introduced new measures to prevent Chinese developers from improperly using the outputs of advanced American AI models (a practice known as distillation) to build rival chatbots, marking the first major U.S. response to Silicon Valley complaints.
The White House Office of Science and Technology Policy will facilitate information sharing, help the industry detect unauthorized model extraction, and collaborate with tech firms to curb such abuses and hold bad actors accountable.
Leading U.S. AI developers including OpenAI, Anthropic, and Google have raised alarms that Chinese rivals like DeepSeek are employing large-scale proxy accounts for "industrial-scale" distillation, bypassing safety guardrails and significantly cutting costs.