China wants artificial intelligence to drive 90 percent of its economic activity by 2030, setting an unusually broad role for the technology across industry, business, and daily services. The plan, however, faces a major constraint: access to advanced AI chips needed to train and operate large-scale models.
The best Huawei AI chip is estimated to deliver performance about five times lower than the leading AI chip currently available in the United States, according to an analysis by the Council on Foreign Relations cited by Kompas Tekno. That gap highlights how hardware capacity may limit the scale and speed of China’s AI expansion.
A National AI Push with a Hardware Problem
Advanced AI chips are central to processing the huge workloads required by modern AI systems. The most advanced technology in this area remains dominated by US companies.
Since 2022, the United States has restricted exports of advanced AI chips to China as competition between the two countries intensifies. The restrictions have pushed China to accelerate domestic semiconductor development and reduce its dependence on US supplies.
China’s homegrown chips have not yet matched the capability of products available in the United States. Strengthening the domestic semiconductor supply chain will therefore be critical to the country’s roadmap through 2030.
AI Is Intended to Reach Far Beyond Chatbots
Beijing has placed the AI agenda within its 15th Five-Year Plan and has expressed an ambition to become a global leader in AI development within the next five years. The technology is being framed not merely as a digital assistant, but as infrastructure for a large share of the economy.
Chinese carmakers have already begun equipping vehicles with AI assistants and intelligent driving features. Technology companies in the country are also introducing AI-powered wearable devices.
Chinese AI video-generation models have raised concerns in the film industry and in Hollywood because of their ability to create highly realistic videos. Humanoid robots are also being positioned for tasks in workplaces and homes.
| Priority Field | Planned Role |
|---|---|
| Humanoid robots | Perform human tasks at work and at home. |
| Workplace AI systems | Automate a range of routine jobs. |
| Low-altitude aviation | Includes flying cars and drone delivery services. |
| Brain-computer interfaces | Use AI to turn neural signals into digital commands. |
| Future industries | Includes nuclear fusion, quantum technology, space, biomanufacturing, and 6G networks. |
Open Models as an Adoption Strategy
China is pursuing ecosystem growth alongside investment by keeping many AI models open source. This allows developers to download, use, and modify models more freely.
The approach differs from many leading US AI models, which are generally closed and require subscriptions. China sees broader model distribution as a way to speed up adoption and expand its software ecosystem.
Kyle Chan, a researcher at the Brookings Institute, said, “China’s open-source approach is designed to drive adoption by giving away AI models for free, building a wider software ecosystem, and then offering paid services for integration and technical support.” The strategy shifts competition beyond model capability toward the ability to build a large user base.
Investment Spans Several Technology Frontiers
China is reportedly preparing investments worth billions of US dollars to accelerate its priority sectors. The agenda includes low-altitude aviation, where flying cars and drone delivery are expected to play a role.
Brain-computer interfaces are another focus because AI can help translate neural signals into digital instructions. Nuclear fusion, quantum technology, space, biomanufacturing, and 6G networks are also included in the future-industry agenda.
The scope of the plan shows how China intends to embed artificial intelligence across physical industries as well as digital services. Yet the effort to make AI a driver of 90 percent of economic activity will depend heavily on whether China can narrow its advanced chip deficit.