# US-China AI Competition and Regulatory Escalation > Live situation record from CLSTR: https://clstr.news/situations/us-ai-competition-with-china > Updated: 2026-08-30T10:00:00.000Z. Sources: 91. Developments: 11. Following the July 25 Genesis Mission funding announcement, the United States paired investment with diplomacy, confirming a senior-level AI dialogue with China and signing the San Francisco Declaration with South Korea. The declaration created a $950 billion framework for joint work on semiconductors, AI-compute infrastructure, robotics and autonomous vehicles, and pledged deeper research collaboration and supply-chain resilience. In late July, the United States announced it will host its first senior-level government-to-government AI dialogue with China in September 2026. In response to escalating tensions, a Reuters investigation and analysis by the Jamestown Foundation revealed that PLA-linked institutions are using model distillation to train domestic defense-grade AI. By extracting outputs from U.S. models like OpenAI’s GPT-3.5 and Anthropic’s Claude, researchers are transferring Western reasoning capabilities into compact “student” models that run on local hardware. Specific applications include Unit 96941 using GPT-3.5 to process sensitive military source code, and the PLA Air Force unveiling an AI-assisted strike-planning system in August 2026. China has countered U.S. concerns by labeling such accusations as “AI hegemonism.” Beijing is intensifying export controls to protect technological sovereignty, considering restrictions on model weights, training data, technical documentation, and software architectures. This includes a de facto export regime for human capital, requiring official permission for highly qualified AI developers from firms like Alibaba and DeepSeek to travel abroad. To counter Western influence, China is pursuing a blueprint to become a “data superpower” by the end of 2028 via the National Data Bureau. Beijing is exporting low-cost AI modules and high-quality datasets to developing nations to address the current dominance of English-language datasets. Experts warn this could allow Beijing to export official narratives into the global knowledge system, noting that Chinese-language responses in models like ChatGPT often show a bias toward Beijing’s views. ## Claims - Reuters examined more than 80 Chinese academic papers and patents documenting the use of model distillation by military researchers. (disputed by 7 sources) - China rejects the accusations, calling U.S. actions double standards and AI hegemonism. (disputed by 6 sources) - The issue has become a flashpoint ahead of U.S.–China talks on AI governance and safety. (corroborated by 9 sources) - Model distillation uses the outputs of a powerful AI system to train a smaller model that can perform similar tasks with fewer resources. (corroborated by 8 sources) - U.S. officials say the practice may violate export‑control rules and intellectual‑property rights. (corroborated by 8 sources) - Chinese military researchers have used U.S.-developed AI models (OpenAI GPT‑3.5 and Anthropic Claude) to train domestic AI systems. (corroborated by 8 sources) - The technique used is model distillation, where outputs of large models train smaller models that run locally. (corroborated by 8 sources) - More than 80 Chinese academic papers and patents document the use of model distillation for defence purposes. (corroborated by 8 sources) - PLA Unit 96941 used GPT‑3.5 to process and summarise military source code, then trained a domestic model for internal network use. (corroborated by 8 sources) - A Chinese defence university used distillation to compress an image‑recognition model for deployment on drones, enabling real‑time navigation and target identification without communications. (corroborated by 8 sources) - Chinese military researchers have used outputs from leading U.S. artificial intelligence models developed by OpenAI and Anthropic to train domestic AI systems. (corroborated by 7 sources) - Chinese military researchers have used outputs from U.S. AI models developed by OpenAI and Anthropic to train domestic defence AI systems. (corroborated by 7 sources) ## Timeline ### 2026-08-30: China advances AI dominance through energy and data infrastructure China is leveraging its energy resources through the ‘Eastern Data and Western Compute’ initiative to secure AI dominance, sparking concerns over the technological arms race with the United States. 2 sources. https://clstr.news/cluster/china-advances-ai-dominance-through-energy-and-data-infrastructure ### 2026-08-25: China establishes 200 AI standards and targets 6G development China has developed nearly 200 AI standards to regulate its growing tech sector and is accelerating 6G and brain-computer interface research as part of its 15th Five-Year Plan. 11 sources. https://clstr.news/cluster/china-develops-nearly-200-key-ai-standards ### 2026-08-17: China seeks to dominate global AI data supply to counter Western influence China aims to become a global AI data superpower by 2028, exporting high-quality datasets to counter Western influence and integrate official narratives into global AI training models. 2 sources. https://clstr.news/cluster/china-seeks-to-dominate-global-ai-data-supply-to-counter-western-influence ### 2026-08-16: China expands export controls to include AI technology China is expanding export controls to include AI model weights, training data, and specialized expertise to safeguard its technological sovereignty and national security. 9 sources. https://clstr.news/cluster/china-expands-export-controls-to-include-ai-technology ### 2026-08-05: Chinese Military Trains AI Systems Using OpenAI and Anthropic Models Reuters found Chinese military‑linked groups using OpenAI and Anthropic model outputs to train domestic AI via model distillation, signaling a push to embed advanced Western AI reasoning into Chinese defense R‑ 2 sources. https://clstr.news/cluster/chinese-military-trains-ai-systems-using-openai-and-anthropic-models ### 2026-08-04: China moves private firms into defence AI as PLA pursues disruptive tech Chinese firm Jingan touts its AI situational‑awareness system as a Palantir rival, while the PLA intensifies a drive for disruptive tech like AI and hypersonics under a military‑civil fusion plan. 3 sources. https://clstr.news/cluster/china-moves-private-firms-into-defence-ai-as-pla-pursues-disruptive-tech ### 2026-08-02: China uses US AI models to train military systems China’s PLA is using U.S. AI models via model distillation to create domestic military AI, including a new AI‑assisted strike‑planning system, despite U.S. export‑control concerns. 10 sources. https://clstr.news/cluster/chinas-military-distills-us-ai-models-for-defense-systems ### 2026-07-31: China's Military Uses US AI Models to Boost Defense Systems China’s military is using U.S. AI models through model distillation to develop domestic defence AI for drones, cyber‑warfare and surveillance, sparking U.S. export‑control concerns. 15 sources. https://clstr.news/cluster/chinese-military-uses-us-ai-models-through-model-distillation ### 2026-07-28: Chinese Military Researchers Use U.S. AI Models via Model Distillation Chinese military researchers have used U.S. AI model outputs via model‑distillation, sparking U.S. concerns over export‑control violations and a diplomatic flashpoint ahead of AI talks. 35 sources. https://clstr.news/cluster/china-and-united-states-unveil-rival-ai-cooperation-blocs ### 2026-07-26: US Leads New AI Diplomatic Initiatives with China and South Korea The US plans a first‑ever AI dialogue with China for September 2026 and signed a $950 bn AI cooperation declaration with South Korea, focusing on safety, technology supply chains and joint research. 2 sources. https://clstr.news/cluster/us-leads-new-ai-diplomatic-initiatives-with-china-and-south-korea ### 2026-07-25: U.S. Announces $5 bn AI‑Centric Science Initiative to Stay Ahead of China The U.S. science office unveiled a $5 bn AI‑focused research push, targeting AI, quantum computing and other tech to keep ahead of China, with bipartisan debate and allied implications. 5 sources. https://clstr.news/cluster/us-announces-5-bn-aicentric-science-initiative-to-stay-ahead-of-china --- Cite as: US-China AI Competition and Regulatory Escalation. CLSTR, https://clstr.news/situations/us-ai-competition-with-china