Chinese technology companies are accelerating investment in AI models, semiconductors and computing capacity as they seek to narrow the gap with U.S. competitors amid tighter U.S. export controls on advanced AI semiconductors to China. Tencent recently signed a five-year, approximately $7 billion computing-resource lease with U.S. cloud provider Oracle, covering multiple facilities in Southeast Asia, with 30% reportedly paid upfront. The agreement gives Tencent access to about 100,000 advanced AI chips unavailable within China, supporting its Hunyuan model and self-operating AI agent tools. Tencent plans to more than double AI spending this year from the previous year; its second-quarter capital expenditure rose 176% year-on-year to 52.8 billion Chinese yuan, while free cash flow turned negative for the first time in about a decade. Alibaba plans at least 380 billion Chinese yuan in AI and cloud infrastructure spending over three years and has unveiled the Zhenwu V900 AI accelerator, alongside plans for a 5 trillion-to-10 trillion-parameter model and an AI full-stack ecosystem. ByteDance raised a $29.6 billion syndicated loan from about 30 banks to fund data centers and model development, while also pursuing a world model. DeepSeek, Z.AI, Moonshot AI and Minimax are likewise expanding, although OpenAI and Anthropic have accused them of unauthorized model distillation. Chinese officials continue to reject calls for slower development, while industry experts describe the U.S.-China race for superintelligence as a prisoner’s dilemma.