Zhipu AI's shares, listed in Hong Kong as Knowledge Atlas Tech Joint Stock (HK:2513), staged a notable rebound on Tuesday, jumping more than 20% to reach as high as HK$698.0 during the session.
The recovery followed a near 23% drop in the previous trading session. That earlier sell-off occurred after Zhipu publicly solicited partners last week to provide additional computing resources intended to help run its artificial intelligence models.
Investor concern over a constrained supply of compute capacity has been accompanied by growing consumer complaints about a deterioration in the performance of the company’s AI models, complaints that market participants linked to the reported lack of computing resources.
Despite the recent volatility, Zhipu’s shares remain trading roughly 400% above their initial public offering level, reflecting continued investor enthusiasm for China’s AI sector.
Zhipu was the first of the cohort dubbed China’s 'AI tigers' to go public. It was followed by peer MiniMax Group Inc (HK:0100), which listed at the start of the year. The broader episode of share price swings has unfolded against additional industry headlines: on Monday, U.S. AI startup Anthropic publicly accused MiniMax and two other named Chinese AI firms - Moonshot and DeepSeek - of appropriating data from Anthropic’s Claude chatbot. According to market reports, those accusations did not materially affect the shares mentioned in this article.
The sequence of events highlights investor focus on two operational constraints for AI companies: access to adequate computing capacity and the implications of data-related disputes. For Zhipu, the immediate market response was a sharp rebound in share price after the company moved to secure external compute partnerships.
Market context and investor response
Market participants reacted to both the company’s public appeal for computing partners and the accompanying consumer reports about model performance. The stock’s volatility underscores how operational bottlenecks and publicized disputes can quickly influence investor sentiment in listed AI companies.