Zhipu stock is surging as U.S. curbs on Anthropic's top AI models boost Chinese rivals

Behind the rally was a Friday directive from the Trump administration requiring Anthropic to cut off access to Fable 5 and Mythos 5 — its two flagship AI models — to all foreign nationals, a category that includes non-citizen workers on Anthropic's own payroll, on grounds of national security. Anthropic said it chose to disable both models for all users to ensure compliance, given the breadth of the restriction.

Coinciding with Washington's announcement, Zhipu said it planned to make GLM-5.2 — its newest open-source large model — publicly available later this week, free of any access restrictions. Zhipu appeared to frame the release as a direct response to Washington's intervention. "Cutting-edge intelligence should not belong to only a few, nor should it be withdrawn at any time," the company said. "It should be open, available, extensible and built to serve every developer."

Wall Street banks sharpened their stances on both companies Monday, according to CNBC. JPMorgan $JPM kept its overweight rating on Knowledge Atlas Technology and pushed its price target up to HK$1,400 from HK$950, even as it cut its view on MiniMax. MiniMax shares still managed a 7.4% gain. In a separate move, Bank of America $BAC opened coverage on both stocks at buy, assigning a HK$1,250 target to Zhipu and HK$500 to MiniMax.

The dynamic has played to the advantage of Chinese developers, whose embrace of open distribution is pulling in enterprise customers who balk at the rising costs of U.S. frontier models. Bank of America analysts argued that as American providers push prices higher, Chinese alternatives are winning over buyers in what the bank describes as the "value-for-money" tier of the global AI market.

The Anthropic restrictions have also renewed debate over the AI talent competition between the U.S. and China. Z-Ben Advisors managing director Peter Alexander put the share of U.S.-based AI engineers with Chinese origins at roughly 40%, a figure that underscores how broadly the new restrictions fall on the very researchers who contributed to building these systems. In a Monday note, he cautioned that the policy could trigger a wave of "brain flight" benefiting Chinese AI firms.

The Anthropic dispute has a broader backdrop. Anthropic sent senior technical staff to Washington over the weekend to negotiate with government officials, and dispatched co-founder Tom Brown and public policy chief Sarah Heck to meet with Commerce Secretary Howard Lutnick and National Cyber Director Sean Cairncross. The company disputes the severity of the security finding underlying the export controls, arguing the level of capability demonstrated by the alleged jailbreak is available from other publicly deployed models, including OpenAI's GPT-5.5.

The shutdown arrives at a delicate moment for Anthropic, which filed a confidential IPO prospectus with the Securities and Exchange Commission earlier this month, disclosing a revenue run rate of $47 billion and a valuation of $965 billion. Since going public in January, Zhipu's share price has multiplied more than ten times over, lifting its total market value to HK$489 billion as of Monday.