Contemporary Amperex Technology Co., Limited (CATL) is building an AI-native research system to accelerate battery materials discovery, Chairman and CEO Robin Zeng said at the 2026 World Laureates Forum at Shanghai Lingang Center.
The system uses 600 TB of data and more than one trillion testing data points. It combines virtual, or dry, experimentation with physical, or wet, testing in a closed-loop process. According to CATL, the approach has enabled researchers to screen more than 1,600 compounds and identify seven promising candidate materials in a short period.
Zeng said battery research involves substantial complexity across multiple scales. Relevant processes can range from femtoseconds to a service life of 25 years, while spatial scales extend from individual atoms to systems measuring tens or hundreds of meters. He noted that current battery energy density remains insufficient for passenger aircraft and other demanding applications.
CATL shipped 661 GWh last year and expects shipments to exceed 1,000 GWh, or 1 TWh, this year. Zeng described this level as the company’s entry into the TWh era and said operating at that scale places greater demands on technology. The company’s research and development activities are supported by more than 60,000 patents and proprietary know-how.
Zeng called for closer cooperation between scientists and the battery industry, including work to establish scientific models in physics, chemistry and electrochemistry that extend beyond current AI models. CATL has focused on translating scientific discoveries into technologies, products and commercial applications for energy generation and use.
He also said progress in sustainable energy requires an environment that respects innovation, supports technology licensing, and protects patents and know-how. He urged entrepreneurs and technology leaders to collaborate in developing such an ecosystem.
Source: CATL press release
