How AI could save Made in Italy fashion sales in China And especially Alibaba's LLM, which is revolutionizing the business of many brands

Three weeks ago, just before the start of the Christmas holidays, the second edition of the SDA Bocconi report titled The contribution of Alibaba's digital marketplaces to the European economy and the sustainable international development of SMEs was presented in Rome, at the Chamber of Deputies. The study analyzes the impact that China's mega e-commerce portals, and especially Tmall Luxury Pavilion, which is the most important in the country, have on the Italian national economy and in particular on fashion. This type of export, which does not concern European or Italian products sold in physical stores but online, is defined as “digital exports”.

The study actually refers to data relating to the three-year period between 2022 and 2024, which represent the most complete and consolidated snapshot available of the situation of Italian exports through Chinese digital channels. The interesting thing, however, is that the data show how these platforms generated, in the six main economies of the European Union — which include Belgium, France, Germany, Italy, the Netherlands and Spain — a total contribution to Gross Domestic Product of 38.6 billion euros, supporting over 143,000 jobs in 2024 alone, both direct and indirect and induced.

In the Italian case in particular, digital exports to China exceeded the threshold of 5.19 billion euros in 2024, a very significant growth despite the consumption crisis that has clearly slowed sales in China itself. And now many fashion brands are waiting, as a test, for the arrival of Chinese New Year 2026, which will run from February 17 to March 3. A moment that could truly confirm the role of Alibaba's platforms as the main driver of international expansion for European brands.

Chinese e-commerce and European exports

The SDA Bocconi report integrates the analysis of direct sales to China with new elements relating to expansion in the markets of the Association of Southeast Asian Nations countries, abbreviated as ASEAN, and to the impact on tourism, given that Alibaba's ecosystem, which includes not only e-commerce but also logistics services, digital marketing and travel platforms, has become over the years a unique synergistic set of services capable of mitigating the typical volatility of traditional trade.

In the three-year period analyzed, amid a general contraction of European exports to China, which recorded a drop of 8.4% over the three years and 6.3% just between 2023 and 2024, sales through Alibaba's digital channels held up very well instead: the overall drop over the three years for the six economies was only 2.4%, while between 2023 and 2024 there was an increase of 7.5%. According to the study, all this is due to more direct sales paths to consumers, to real-time insights on market trends and to a rapid response capacity to the variations of an increasingly volatile demand that physical retail apparently cannot intercept at the same speed.

This type of stability is the result of a facilitated commercial policy, so to speak, which has helped, for example, many Italian brands (fashion alone accounts for 67% of sales, followed by beauty) to add half a billion euros in growth to their businesses in 2024 alone. The policy consists in the fact that platforms like Tmall Global allow small and medium-sized enterprises to access the vast Chinese market without the obligation to establish a local legal entity or to maintain warehouses in the country, greatly reducing operating costs and entry barriers that might otherwise discourage smaller entities.

But what does this innovation consist of?

Certainly a very important part of this innovation comes from the ramification of digital ecosystems that go far beyond simple e-commerce. With its 200 brands in the lineup, Tmall Luxury Pavilion has begun to offer immersive and highly personalized shopping experiences, designed especially for Millennials and Gen Z in the country, who are attracted thanks to augmented and virtual reality features that allow virtual try-ons for jewelry, watches, clothing and accessories, as well as interactive 3D visualizations and virtual showrooms that faithfully replicate real physical boutiques, for example inspired by iconic places like Place Vendôme in Paris. As early as 2022, for example, Jing Daily reported that sales of luxury goods with 3D and AR functions had recorded double-digit growth on an annual basis.

It's not just try-ons, though. The immersive digital experiences offered by Tmall include AR fashion shows, virtual avatars, digital art galleries and “exclusive products” (it would be better to talk about subscriptions) like the Meta Pass that guarantees priority access to limited-edition products. Among the other services available there are then in-depth personalizations through video consultations, entire spaces dedicated to post-purchase maintenance on how to adjust watches and bracelets, how to clean jewelry and shoes and which products to use. But the real innovation is a new AI called Qwen.

When AI becomes a business

Alibaba's proprietary large language model, called Tongyi Qianwen or familiarly Qwen, is perhaps the most public innovation that is talked about the least in the mainstream fashion sector. To date, Alibaba's Qwen model (now the third updated iteration of the model is used) provides personalized recommendations based on local trends, purchasing behavior and special occasions; it offers digital assistants for gift suggestions and even generates its own content autonomously.

Alibaba is investing heavily in this AI model that can integrate cloud, e-commerce and artificial intelligence for applications dedicated to customers but that can also be sold to companies. According to the company itself, the AI-based B2B business market is already worth 30 billion dollars globally. We are talking here about the features that Alibaba itself offers (not without some controversy) to connect brand creators around the world with Chinese factory suppliers. It is precisely the AIs that make this entire logistical and research process much smoother.

There are over 300 open-source generative AI models in the Qwen family and 228 Qwen model launches in 2025, including six new Qwen3 variants in a single week. AI, to quote Alibaba itself, «is able to extract meaning from unstructured inputs, including product sketches, engineering designs, documents, certificates, production capacities and factory track records, and match them to relevant production capacities» unlocking «the vast network of specialized, customized or regionally focused suppliers — often high-potential SMEs — whose expertise remains invisible in traditional keyword-based search models. By bringing these suppliers to light, AI Mode can help SMEs improve in terms of product innovation and supply chain resilience».

In the cloud, Qwen is integrated into Alibaba Cloud for consumer analysis, supply chain optimization and creation of customized apps, like those used by LVMH to obtain insights and tailored services. Other applications include fraud detection, intelligent logistics, generated content such as AI avatars in live streaming, and tools for merchants such as real-time translation and product copy generation, with a new Qwen Consumer Business Group created in December 2025 to push consumer use of Qwen models and a free app based on the most advanced version of Qwen LLM available in China as a mobile app and website since November 2025.

What does the future of e-commerce in China hold?

For this year, Alibaba aims to make AI self-sustaining in e-commerce. According to several analysts, the goal is to achieve triple-digit growth in AI-related products and massive adoption among so-called merchants, i.e. individuals on the site to do business in any form and not to buy privately, including the adoption of Qwen3 by 290,000 companies and integration into frameworks like Apple MLX as explained by the company itself.

Morgan Stanley predicts that by 2026, up to 40% of B2B transactions could involve AI-agent-to-AI-agent negotiations, with AI commerce potentially growing between 3 trillion and 5 trillion dollars annually by 2030. This is a great advantage for Tmall, which has gotten ahead and could indirectly benefit from these developments, continuing to offer customers premium immersive experiences, while cultivating an overall ecosystem increasingly dominated by AI that make commerce within the “Alibaba galaxy” progressively more efficient, whose cloud division could exceed 30% of total revenue by 2026.

Takeaways

- Three weeks ago, at the Chamber of Deputies in Rome, the second edition of the SDA Bocconi report on the contribution of Alibaba's digital marketplaces to the European economy was presented, with a focus on the positive impact on Italy and the fashion sector through digital exports.

- The study, based on data from the 2022-2024 three-year period, highlights how these platforms generated 38.6 billion euros in GDP across the six main EU economies, supporting over 143,000 jobs in 2024 alone, while Italy recorded digital exports to China exceeding 5.19 billion euros despite the consumption crisis.

- In a context of general decline in European exports to China, Alibaba's channels demonstrated greater resilience, thanks to more direct sales, real-time insights, and facilitating policies such as Tmall Global that reduce barriers for SMEs. Tmall Luxury Pavilion, with over 200 brands, offers immersive experiences through augmented reality, virtual try-ons, 3D showrooms, and advanced personalizations, primarily attracting Millennials and Gen Z.

- The key innovation is represented by Alibaba's Qwen AI model, which provides personalized recommendations, digital assistants, and optimizations for supply chains and merchants, with hundreds of variants launched in 2025.

- The upcoming Chinese New Year 2026 could represent a decisive moment to confirm the role of these platforms as a growth engine for European brands, within an ecosystem increasingly dominated by agentic artificial intelligence.