According to a report by Yonhap on October 19, the Korea Investment Corporation (KIC), which manages assets worth 339 trillion Korean won (approximately 1.7 trillion yuan), is using Chinese AI models for tasks such as processing internal document summaries, translation, and data search, which has drawn attention.
According to data submitted by KIC to the House Committee on Finance and Economic Planning on the 19th, its AI system is equipped with four inference models. The language model is Alibaba’s Qwen3, while the embedding model for retrieval is Beijing Institute of Artificial Intelligence (BAAI)’s BGE-M3. The other two models are Meta Corporation’s Llama 3 and Google’s Gemma3, which are used for blocking harmful content and processing images.
In other words, everything from processing summaries and internal documents, to managing rules within groupware systems and data for congressional business management systems, as well as compiling weekly reports—all these tasks are handled using Chinese-made reasoning models.
It is worth mentioning that the U.S. Department of Commerce listed BAAI on the Entity List on March 25th last year, and KIC began using BGE-M3 in its internal AI system on December 15th of the same year.
Regarding the actions of the US side, BAAI responded at that time by stating that it was shocked by the inclusion of private non-profit scientific research institutions in the entity list. It strongly opposed this wrong decision, which had no factual basis, and demanded that the relevant US departments withdraw it.
Regarding the reasons for using Chinese AI models, KIC explains that after evaluating factors such as accuracy, understanding ability, reasoning ability, business compatibility, and security, Qwen3 has advantages in terms of cost-effectiveness, graphics card memory capacity, and suitability for various contexts, especially in the Asian market.
The Chinese Ministry of Foreign Affairs previously stated that China opposes the use of ideology as a dividing line and technological blockades. It is committed to balancing development with security. China is willing to share experiences with all parties and enhance the level of global artificial intelligence innovation and development.
