China AI Stocks Slide $33 Billion as US Labs Push Frontier Model Restrictions, Z.AI and MiniMax Down Over 30 Percent in September

What Happened Chinese AI developers Z.AI Co. and MiniMax Group Inc. saw their shares plunge more than 30 percent each in September 2026, wiping $33 billion from their combined market value, according to Bloomberg on September 17. The selloff intensified as Anthropic PBC and other US developers called on policymakers to restrict Chinese access to frontier models, accusing some Chinese rivals of using American models without approval to train their own through a practice known as distillation. The push comes as investors scrutinize fierce competition and cash burn rates among Chinese AI firms. China's Ministry of State Security head Chen Yixin argued in an article published September 15 that AI poses threats to China's political and ideological security and could be weaponized, singling out Anthropic's and OpenAI's models as potential cybersecurity threats. China rejected US calls to curb AI development as fearmongering and part of an American Cold War playbook, according to multiple outlets on September 17. Why It Matters The dual collapse in Chinese AI valuations and US advocacy for tighter export controls signals that AI geopolitics is now a capital allocation and national security problem, not just a regulatory debate. If US labs secure distillation restrictions, Chinese developers will face a double squeeze: capital markets pricing in weaker competitive positioning while technical access to frontier models narrows precisely as they attempt to close the capability gap. For investors, the $33 billion wipeout is a preview of valuation risk when geopolitical rivalry overrides market fundamentals. For defence and intelligence planners, the distillation accusation confirms that model leakage is a live strategic vulnerability, not a theoretical one.