China Integrated Circuit Industry Investment Fund (the “Big Fund”), the country’s largest state-backed semiconductor investment vehicle, is negotiating to lead a major external funding round for artificial intelligence startup DeepSeek. Sources indicate that, if completed, the deal could value DeepSeek at up to $45 billion.
The Big Fund would act as the lead investor, with Tencent also expected to participate in the financing, although the final investor lineup has not been settled. At the start of the planned round the company’s estimated valuation was about $20 billion; that figure has risen rapidly during negotiations to the current potential $45 billion level.
Liang Wenfeng, the Hangzhou-based founder who is a billionaire, may also contribute to the round. Corporate filings show that through personal holdings and affiliated entities he controls 89.5 percent of DeepSeek.
Strategic significance of the Big Fund’s involvement
The Big Fund was established by Beijing to strengthen China’s technological self-reliance in the face of U.S. export curbs, with a primary focus on investments in semiconductor equipment and materials. Its portfolio includes key domestic chipmakers such as Semiconductor Manufacturing International Corporation (SMIC) and Yangtze Memory Technologies Corp. Until now, the fund has not publicly backed any other Chinese companies developing large language models.
The planned investment takes on added strategic weight because DeepSeek’s latest V4 model has been optimized for the Huawei Ascend 950PR processors. Sales of Huawei’s MI chips have risen sharply this year, and Huawei has overtaken Nvidia in the Chinese market as advanced U.S.-made products remain restricted. Beijing expects closer collaboration between chip manufacturers and model developers to foster an independent Chinese ecosystem that can sustain competitiveness despite sanctions.
Nvidia’s CEO Jensen Huang has warned that such an ecosystem could threaten U.S. global technological dominance; in a recent interview he said that the day DeepSeek first runs on Huawei hardware would be a dire outcome for the U.S.
Technology potential and transparency concerns
Investors appear willing to bet on DeepSeek’s technical prospects despite the startup not yet focusing on commercial product sales at scale. DeepSeek attracted international attention in January 2025 when it released R1, an open-source large language model. The company claimed R1 was produced at a fraction of the development cost of comparable systems and performs similarly on typical chatbot tasks.
However, several U.S. analysts and business figures have questioned the veracity of the company’s data and its transparency; Elon Musk publicly raised such concerns last year. DeepSeek, the Big Fund and Tencent did not respond to requests for comment from the Financial Times.
Related items: AI definition and media use
The source article included a brief explanation of artificial intelligence — programs or systems that can respond or behave in human-like ways without direct human intervention, capable of reasoning, learning and creativity — and described chatbots as computer programs able to carry out human-like conversations. It also noted that Yahoo Finance used ChatGPT to explore which financial skills matter most during economic uncertainty.
Further clarity on the size and participants in the financing, and any formal statements from the companies and the Big Fund, will be needed to confirm the final terms and strategic implications.
