A fresh wave of competition is rippling through China’s AI landscape after a key personnel move involving Alibaba’s Qwen large language model team.
On March 5, 2026, Alibaba CEO Eddie Wu formally approved the resignation of Jun-Yang Lin, described as a core member of Alibaba’s Qwen team, according to a report from TechWeb. The decision was communicated through an internal letter addressed to Tongyi Lab, Alibaba’s research unit closely associated with its generative AI and large language model efforts.
Alongside the departure, Alibaba announced a structural response aimed at keeping its foundation model roadmap moving fast. The company is setting up a new “foundation model support group,” designed to coordinate internal resources more effectively and reinforce future development of large language models. In practical terms, that kind of support group typically signals a push to streamline execution—aligning research, engineering, product, and computing resources so model training, iteration, and deployment can happen faster and with fewer bottlenecks.
The timing is notable because the race to build stronger foundation models has intensified across the industry. When high-profile talent shifts occur inside top AI labs, they often trigger broader ripple effects: competitors may accelerate recruitment, internal teams may reorganize to protect momentum, and companies may double down on infrastructure and coordination to maintain a steady pace of model upgrades.
For Alibaba, the message is clear: even amid leadership changes, Qwen and Tongyi remain strategic priorities. By adding a dedicated foundation model support group, the company appears to be reinforcing its long-term commitment to large language model development—and signaling that it intends to stay competitive as the AI talent war and model arms race continue to heat up.
