Alibaba (09988) will provide a generative AI basic model solution to Japanese companies.

On March 5, Alibaba Group (09988) announced that its subsidiary Alibaba Cloud will promote the use of generative AI (artificial intelligence) basic models for businesses in Japan and collaborate with Japanese consulting and system development companies. At a press conference held on the same day, the Japan head expressed, "We aim to increase the number of projects using our company's models to over 1000 within 3 years." Alibaba Cloud has basic models such as "Qwen" for natural language processing and "Wan" for generating images and videos. The Qwen 2.5 released in September 2024 is also proficient in programming, mathematics, generating long texts, and has high accuracy in Japanese language processing. So far, the application cases of Qwen in Japan have been limited to some emerging companies for fine-tuning. The Country Manager of Alibaba Cloud Japan Service, expressed, "Previously, there was no systematic exploration of Japanese companies' needs and development." Alibaba Cloud Japan Service will work with consulting and system development companies to adjust and customize models according to the needs of Japanese companies, as well as collaborate on the development of AI applications. Negotiations have begun with several partner companies and proposals will be officially presented by 2025. The Japan head mentioned, "We will prioritize presenting proposals to IT, manufacturing companies, and the retail industry. We plan to introduce successful cases from China to Japan." Although these AI basic models are developed by Chinese companies, the Japan head emphasized their security, stating "In Japan, they will be provided in the form of a private cloud, with data staying within Japan's data centers. Security will also be ensured." This article is reprinted from the WeChat public account "Nikkei Chinese website," GMTEight Editor: Xu Wenqiang.