The 2025 Government Work report proposed that we should continue to promote the "artificial intelligence +" action and cultivate new growth points such as digital trade. The implementation of the "artificial intelligence +" action will promote the deep integration of artificial intelligence and technical resources in various industries, and promote cross-field technical collaboration and innovation. In Smart Trade - How Artificial Intelligence and Trade Activities Shape Both Ways, the World Trade Organization proposes that artificial intelligence will reshape the future of international trade. In the future, digital trade will be more dependent on the efficient circulation of data elements and the deep application of intelligent technologies. Artificial intelligence will also enable the intelligent development of digital trade with effective technologies, affect the technical basis and service model of global digital trade, and bring new opportunities and new changes to industrial development.
Promote innovation in digital trade products and services
Artificial intelligence will trigger a paradigm change in digital trade, promote the development of digital trade to "intelligence-driven", enable technological innovation in the field of digital trade, expand ecological scenarios, etc., and improve the efficiency of cross-border services. In the field of digital product trade, taking digital audio-visual products as an example, previous digital content creation such as novel creation, film and television script writing, music creation, etc., mainly rely on the author's personal creativity and experience, with the natural language processing ability and creative generation ability of artificial intelligence technology, the author can get more creative inspiration and assistance. Artificial intelligence enables the creation of digital content, not only improves the creation efficiency, but also improves the diversity of digital content, and will effectively enhance the supply capacity of digital products and services.
In the field of digital technology and service trade, artificial intelligence will greatly enhance the personalized and intelligent supply capacity of digital services. Taking online education as an example, artificial intelligence can understand students' learning status and needs through intelligent analysis of students' learning data, including learning progress, answers, learning habits, etc., to form personalized learning plans. At the same time, AI can also be used as an intelligent tutor to achieve online tutoring 24 hours a day and answer students' questions at any time. This intelligent, efficient and individualized online education service will attract students from all over the world to a greater extent and promote the realization and expansion of cross-border digital education services. In the process of medical service, artificial intelligence can quickly and accurately identify abnormal conditions such as tumors and lesions through systematic analysis of patients' medical record data and medical image data, assist doctors in diagnosis, and improve the accuracy of diagnosis of complex cases. Combined with information and communication technology, doctors can consult with patients remotely, which not only improves the accessibility of medical services, but also realizes the optimal allocation of medical resources and promotes cross-border digital medical services. Up to now, Internet medical and health service platforms such as Chunyu Doctor, Jingdong Health, Ping An Health, Tencent Health, and Weemai have been connected to DeepSeek.
In the field of data trade, open source AI large models can promote the transformation of data elements into "dynamic assets" through technological innovation and ecological collaboration, mining the value of data elements to a greater extent, and activating the development of cross-border data elements market. With its hundred-billion-level parameter scale, multi-modal processing capability and efficient reasoning performance, AI large models can enable the whole chain for data trade. For example, through natural language processing technology, data needs can be accurately analyzed, and data products can be quickly matched with market needs. Automated identification of data risk points with intelligent compliance systems; Based on in-depth analysis, dynamic pricing models can be generated to provide efficient solutions for data trade. For example, the Shenzhen Data Exchange has significantly improved the accuracy and richness of the generated content by integrating the DeepSeek R1 model with the RAG solution to build an exclusive private knowledge base, while generating a dynamic pricing model based on in-depth analysis to provide pricing and AI compliance solutions for exploring cross-border data flows in the Guangzhou-Hong Kong-Macao Greater Bay Area.
Promote the transformation of platform enterprises into "intelligent service providers"
Artificial intelligence technology forms digital technology solutions through algorithm optimization and open source ecology, helps e-commerce enterprises overcome development bottlenecks such as language barriers, low operational efficiency, and information asymmetry, enables cross-border e-commerce platforms to transform from "platform intermediaries" to "intelligent service providers", and has a far-reaching and positive impact on the development model and efficiency improvement of digital trade. First, it can realize intelligent matching engine. Enterprises can quickly integrate commodity data, consumer behavior, social media public opinion and other information through multi-modal large models to achieve real-time demand forecasting and supply chain linkage. According to the Coresight Research report, SHEIN uses AI to generate design draft and link flexible supply chain, the shelf life of new products is shortened to 7 days, and the SKU turnover rate is 98%. Lianlian digital integrates DeepSeek (LLM) large language model capabilities in the LOOP AI business application, and generates more accurate, vivid and attractive copy descriptions for cross-border businesses through DeepSeek's natural language accurate interaction and scene perception. Significantly improve the exposure and sales conversion of goods and services on platforms such as Amazon and major search engines. Second, it can realize intelligent service. The use of large models can promote cross-border e-commerce intelligent product selection and supply chain collaboration. Cross-border e-commerce platforms can generate a regional explosive list by analyzing overseas social media hot words and Google Trends data through NLP. Blockchain +IoT technology can realize the full traceability of cross-border logistics. According to DHL statistics, the AI response speed of abnormal events such as customs detentions is 40 times faster than that of human workers. Ali International has also begun to conduct AI tests on more than 40 key scenarios of cross-border e-commerce, covering the full link of cross-border e-commerce, and landing AI applications on AliExpress, Lazada, Trendyol and other platforms, covering 500,000 merchants and 100 million products.
Help smes participate more in global digital trade competition
DeepSeek and other high-performance open source AI models independently developed by China, with low cost, high efficiency and multi-modal capabilities, may break through the "Matthew effect" and help micro, small and medium-sized enterprises to participate in global digital trade competition. On the one hand, it breaks through language and cultural barriers and effectively expands the global market. Language barrier is a major problem for smes to achieve cross-border trade. According to Ali Research, 30% of cross-border e-commerce disputes are due to communication misunderstandings caused by language barriers. The multi-language large model can support real-time translation and sentiment analysis of hundreds of languages, effectively solve the problem of cultural context differences, and make it easy for smes to carry out cross-border language communication. Yiwu Commodity City merchants generate product promotion videos in English, Arabic and other languages through AI to achieve multi-role and multi-language video delivery. On the other hand, smart service upgrades are achieved at a low cost. Large models such as DeepSeek, with their lightweight deployment requirements and high-performance reasoning capabilities, can lower the technical threshold to a minimalist configuration of "business expert + algorithm expert + engineering expert". DeepSeek white paper shows that DeepSeek uses the "distillation training +MoE architecture" hybrid expert model, the training cost is only 1/10 of the international mainstream model, can achieve AI deployment at a lower cost, and help more small and medium-sized enterprises quickly build industry agents. In the future, AI technology may lower the threshold for smes to participate in global trade to a greater extent, and promote the formation of a "digital multipolarization" order.