Published on Dec 17, 2025

Yanfeng Automotive signed a cooperation agreement with Alibaba Cloud in early December covering a “full-stack AI partnership”. According to Chinese technology portal Zhongguancun Zaixian(in Chinese), the wide-ranging collaboration spans multiple stages of the two companies’ value chains, from the use of AI in intelligent manufacturing at the supplier to its integration into smart-cockpit products, and the migration of corporate IT systems to the cloud.

With annual revenues of more than US$15 billion from digital cockpits and other interior products, the Tier-1 supplier is joining a growing list of automotive industry customers that Alibaba has so far secured for its vertically integrated AI solutions.

Among China’s automakers, the internet group has been highly successful for several years with its AI business, which is anchored within the group’s Cloud Intelligence Group segment. Suppliers are now becoming increasingly important for further growth.

“Alibaba Cloud now serves all major automotive customers. We are fully confident in enabling the industry to accelerate into the new AI era,” said Li Qiang, Vice President of Alibaba Public Cloud and General Manager of its automotive business, at the contract signing.

At the product level, the two companies said they plan to jointly develop “next-generation AI solutions for intelligent cockpits” based on Alibaba’s large language model Qwen.

In manufacturing, Yanfeng also intends to deploy Alibaba Cloud’s full-stack cloud technologies to improve efficiency and quality. The stated goal is to optimise end-to-end production management. "‘AI plus manufacturing’ would enable significantly faster development cycles", the companies said. In the context of China’s automotive industry, “full-stack AI” typically means that a provider can offer cloud and compute infrastructure, data pipelines, families of large language models, tools for training and deployment, and industry-ready solutions for automakers and suppliers as one coherent, integrated system.

Put simply, such comprehensive AI offerings function as a kind of ‘Swiss Army knife’, with applications spanning the customer’s entire value chain.

Alibaba’s various cloud-based models are expected to be used at Yanfeng not only in R&D and production, but also in management and operations, to build what the companies describe as an “integrated, scalable digital organisational structure”. Chinese trade media report that Yanfeng, which operates at 240 locations worldwide, has already migrated its entire IT architecture outside China to Alibaba Cloud.

In parts, the extensive list of cooperation elements reads like a summary of Alibaba’s automotive strategy, which, similar to Google – is based on an highly capital-intensive approach aimed at participating long-term in the transformation of OEMs and suppliers in the automotive industry.

At its developer conference in Hangzhou in September, Alibaba once again clearly articulated its ambition to become a “globally leading full-stack AI provider”. The company is strongly focused on B2B business, not only in the automotive sector but also in industries such as advanced manufacturing more broadly, healthcare, robotics and financial services. More than a 250.000 companies have already begun using Qwen since the model was launched in April 2023, Alibaba Cloud writes on its website.

Despite volatility in international semiconductor markets, the company continues to invest heavily in AI infrastructure, seeking – much like Google in the United States – to strengthen its position as a domestic “full-stack” market leader before competing globally with Google.

In February this year, Alibaba announced plans to invest an additional RMB 380 billion in artificial intelligence and cloud architecture over the next three years. That equates to roughly US$54 billion and exceeds what China’s leading cloud provider has spent on these areas over the past decade.

This strategy of investing such large sums in AI infrastructure will have implications for both China’s and the global automotive industry in the years ahead. It enables automakers and their suppliers to accelerate AI adoption, which in turn significantly eases the commercialisation of data-intensive technologies such as autonomous and connected driving.

Put differently, scaling Alibaba’s AI infrastructure will accelerate the scaling of autonomous driving in China. Emerging mobility industries such as robotaxis and unmanned aerial vehicles are also likely to receive a boost.

For Chinese automakers such as GAC, which on November 14 also agreed to expand its cooperation with Alibaba Cloud on full-a AI, the company’s massive AI investment bet acts as a powerful accelerator for their strategic plans. Under the new agreement, the GAC Group will “use Alibaba Cloud’s high-performance computing solutions and elastic architecture to support a wide range of core businesses,” Chinese automotive portal Gasgoo reported.

“This includes migrating key internal systems to the cloud, scaling the development of driver-assistance features and intelligent cockpit solutions, expanding connected-vehicle services, and supporting GAC Group’s global operations,” the portal said.

How does Alibaba intend to make money from this? By combining its own powerful infrastructure with cloud services, Alibaba aims to lock automotive enterprise customers into its ecosystem over the long term and monetise AI through recurring platform revenues.

In addition to the Qwen model family, Alibaba therefore offers multiple variants and specialised derivatives, including multimodal models as well as versions optimised for automotive use cases such as intelligent cockpits, vehicle software and manufacturing processes.

This strategy is beginning to pay off. Revenues at Alibaba Cloud are rising, and for the current fourth quarter Goldman Sachs and other banks are forecasting growth of around 30 percent or more for the Chinese AI business.

Alibaba is becoming increasingly attractive as an AI service provider for automakers and automotive suppliers for several reasons. Independent benchmarking firms report that the performance of Alibaba’s AI models has steadily improved over the past year.

At the same time, Alibaba CEO Eddie Wu Yongming has repeatedly emphasised his commitment to maintaining the open-source strategy of the Qwen model family. The more developers use Alibaba’s AI models and tools, the faster the ecosystem grows – and the larger the base of applications and experience from which enterprise customers such as Yanfeng, GAC, Li Auto or BMW can benefit. Thanks to this combination of improving model quality and an open-source strategy, Qwen is gradually emerging as one of the preferred foundation models for developers in the automotive industry.

While OpenAI continues to pursue direct monetisation of closed models – and Meta founder Mark Zuckerberg is reportedly once again considering a more closed approach – Alibaba is opting for open distribution and is seeking to strategically position itself across as many layers of the automotive value chain as possible.

If you want, I can next
– tighten this for FT / Economist length, or
– adapt tone and framing for US vs European editors, or
– flag passages that might trigger legal or sourcing questions in Anglo-Saxon newsrooms.