Alibaba creates CEO-led task force to sharpen its AI model push

A CEO-level bet on foundational models

Alibaba Group (阿里巴巴集团) has created a new company-wide task force focused on artificial intelligence model development, led directly by CEO Eddie Wu, the South China Morning Post reported. The initiative aims to accelerate progress on Alibaba’s in-house foundation models—best known under the Tongyi Qianwen (通义千问) and Qwen brands—and to tighten integration of core AI capabilities across e-commerce, cloud, and enterprise services. Why elevate it to the chief executive’s desk? Centralized oversight could break internal silos and speed the long path from research breakthroughs to revenue-generating products.

Context for Western readers

Alibaba is China’s largest e-commerce operator and a top cloud provider through Alibaba Cloud (阿里云), which has positioned its models for both consumer-facing assistants and enterprise use cases such as code generation, search, advertising, and customer service. The reported task force follows a period of strategic recalibration, including Alibaba’s broader restructuring and a renewed emphasis on AI after shelving plans to spin off Alibaba Cloud in 2023. In China’s platform economy, owning a competitive foundation model is increasingly seen as table stakes rather than a moonshot.

Geopolitics and compute constraints

The move comes amid tightening U.S. export controls on advanced chips used to train large models, which have limited access to Nvidia’s top-tier GPUs in China and complicated build-out plans across the sector. Chinese firms have turned to efficiency techniques and domestically produced accelerators from players like Huawei (华为) to bridge the gap. Within China, regulators have also rolled out rules for generative AI services, shaping how models are trained, evaluated, and deployed at scale. Against this backdrop, consolidating model development under a CEO-led task force signals priority—and urgency.

Competitive landscape and what to watch

Alibaba’s rivals are pressing hard: Baidu (百度) is commercializing its Ernie family, Tencent (腾讯) is rolling out Hunyuan, and ByteDance (字节跳动) is pushing its Doubao assistant across consumer apps. Success will hinge on model quality, inference cost, and rapid productization inside Alibaba’s shopping, logistics, and payments ecosystems. Watch for signs of deeper alignment between Alibaba Cloud and commerce units, faster release cadences for Qwen/Tongyi models, and new partnerships with enterprises seeking China-based AI infrastructure. Reportedly, the task force’s mandate is to make those pieces fit—and move faster.