In late July, DeepSeek, an AI startup based in Hangzhou, suspended its second fundraising round, at least temporarily, informing potential investors verbally that they shouldn’t anticipate any signatures in the upcoming days. The suspension followed the circulation online of a transcript of what appeared to be a private investor meeting from the company’s initial funding round, in which founder Liang Wenfeng was open about DeepSeek’s dependence on Nvidia chips and admitted that China still lags behind the US in some areas of AI sophistication. Bloomberg has not confirmed the transcript’s veracity. According to people with knowledge of the situation, Liang’s dissatisfaction with how private discussions became a public spectacle was a contributing factor in the suspension. DeepSeek’s record-breaking June fundraising round brought in about $7 billion, with backers including Tencent Holdings and battery giant CATL. The company then started a second round with a pre-money valuation of at least 480 billion yuan, or roughly $74 billion, and a target of at least 10 billion yuan in additional capital. Last year, DeepSeek unveiled a model that proved a state-of-the-art, effective AI platform could be constructed with less computing power than Silicon Valley behemoths were using, an accomplishment especially difficult given US export restrictions on cutting-edge chips.