Federation of Industry and Commerce: Taiwan should transform from an AI hardware supplier to an internationally trusted system partner
Amid the AI competition between the United States and China, the Federation of Industry and Commerce recommends that Taiwan develop model security testing, information security protection and verification systems
Quick Look
- The United States and China have agreed to engage in dialogue on AI security.
- The Federation of Industry and Commerce recommends that Taiwan, under the cooperation between the United States and China in AI, should develop model safety testing, information security protection and international verification systems in addition to hardware production, and transform from a hardware supplier to an indispensable system partner for the global AI industry.
AI-generated summary
Why It Matters
The United States and China agreed to conduct exchanges and dialogues on AI security and risks after the second meeting between Trump and Xi. The two parties are expected to establish an AI accident reporting and crisis communication channel.
The United States and China agreed to conduct exchanges and dialogues on the safety and risks of AI (artificial intelligence) after the second meeting between Trump and Xi, and experts expect to establish an AI accident notification and crisis communication channel. Talking about Taiwan's positioning in the AI competition situation between the United States and China, the Federation of Industry and Commerce suggested that in addition to producing hardware, Taiwan should develop model safety testing, information security protection, product usage verification and overseas maintenance services in the future, and establish an internationally trusted verification and recording system. Only in this way can Taiwan further become an indispensable system partner for the global AI industry from a hardware supplier.
Huang Jianqun, Director of the Mainland Affairs Bureau of the Federation of Industry and Commerce, wrote an article in the latest issue of the Industry Magazine of the Federation of Industry and Commerce. It is unlikely that the second meeting between Sichuan and Xi will eliminate the AI competition between the United States and China. A more practical goal is to delineate different risk levels. Through clear risk classification, we can avoid excessive expansion of the scope of national security controls and affect normal scientific and technological exchanges and commercial activities. Among them, advanced chips, military AI, core model technology and sensitive data will still be strictly controlled, while mature chips, general business models and low-risk applications can retain trading space. As for areas where both parties may be harmed, such as cyber attacks, deepfake fraud, biosecurity and military misjudgment, incident reporting and crisis communication channels are required.
Huang Jianqun pointed out that if a fixed AI dialogue can be launched after the second meeting between Trump and Xi, the United States and China will begin to establish clearer competition rules. The summit can only determine the general direction. What really affects the operation of enterprises is how to regulate product licensing, safety testing, use verification and accident reporting after the meeting. Areas involving advanced chips, military applications, and sensitive data will still be strictly restricted by both parties. However, general commercial clouds, industrial models, and research applications may maintain exchanges as long as they comply with security regulations. The future U.S.-China AI relationship will adopt different levels of control based on technical risks, but it will not be completely divided into two camps that do not interact with each other like during the Cold War.
Huang Jianqun pointed out that the United States’ advantages lie in sufficient funds, leadership in advanced chips and cloud platforms, and the ability to jointly formulate rules with its allies. China is good at reducing the cost of using models and quickly introducing AI into manufacturing, logistics, automobiles, and urban governance. The key to the future AI competition between the United States and China, in addition to model efficiency, also depends on which side can establish stable computing power and energy supply, improve corporate productivity, and get more countries to accept its technical standards.
Talking about Taiwan's position in the future multi-track AI order, Huang Jianqun said that Taiwan has advantages in wafer manufacturing, advanced packaging, servers and supply chain management. Most of the world's advanced AI chips are still manufactured by Taiwanese companies. In the future, in addition to producing hardware, Taiwan should also develop model safety testing, information security protection, stable power supply, product usage verification and overseas maintenance services, and establish an internationally trusted verification and record system. Only in this way can Taiwan further become an indispensable system partner for the global AI industry from a hardware supplier.
Open Questions
- What are the specific implementation details of the US-China AI risk classification?
- How can Taiwanese companies implement the verification system of international trust?







