Alibaba’s Qwen Gains Ground as US Export Controls Block Claude in Hong Kong

Published by James Harris on June 19, 2026June 19, 2026

Alibaba chairman Joe Tsai used a Paris tech conference this week to frame AI as a $50 trillion market opportunity and commit the company to investing across the entire stack: chips, cloud infrastructure, its Qwen model family, and consumer-facing applications. The breadth of that bet is the point. Tsai said it is still too early to know where AI profits will concentrate, which is why Alibaba is not picking a lane.

That positioning matters more now because the competitive landscape is shifting in Alibaba’s favor through regulatory accident rather than product merit. JPMorgan Chase has restricted Hong Kong staff from using Anthropic’s Claude models, following a similar move by Goldman Sachs, with the limitation tied to language in Anthropic’s licensing agreement and tightening US export controls. Anthropic has already blocked foreign nationals from accessing two of its models under a US export-control order. Hong Kong has historically operated under different rules than mainland China, with access to American AI products determined largely by the companies themselves. That buffer is now eroding, and it creates a vacuum that Alibaba’s Qwen model suite is structurally positioned to fill, regardless of whether Qwen is the technically superior product.

Prediction markets are not buying Alibaba’s pitch: Polymarket gives it less than 1% odds of having the best AI model by end of June, against Anthropic’s 95%, and Kalshi puts Alibaba below 1% for best coding model by end of 2026.

Agents on the Ground

The enterprise and consumer pushes are already underway. Alibaba.com launched Accio Work in Malaysia, an AI agent suite aimed at small and medium-sized businesses that can handle sourcing, product listings, marketing, and store management autonomously rather than just answering queries. Alibaba Cloud, meanwhile, launched a new cloud region in France, its third in Europe, with agentic AI services for European clients planned later this year. At home, China’s Ministry of Commerce issued a 17-stage strategy across eight ministries to integrate AI into retail, smart devices, robotics, and consumer spending, explicitly targeting millions of homes and businesses. The domestic policy environment and the international access restrictions are compressing in the same direction at the same time.

The export control pressure on US AI firms in Asia is accelerating a bifurcation that was already underway, and financial institutions operating across jurisdictions are now being forced to make explicit choices about which AI ecosystem they rely on. For Alibaba, the $50 trillion TAM argument is secondary to a simpler near-term reality: its competitors are being partially locked out of the markets where Alibaba is already building infrastructure.