China is advancing plans to enable cross-border data trading as demand for artificial intelligence grows, with a new agreement between the Shenzhen Data Exchange (SZDEX) and Malaysia's Zetrix AI marking a key step. The memorandum of understanding, signed on May 22, aims to explore connections between China's and Southeast Asia's developing data markets. Data is increasingly seen as a critical resource for AI systems, which require localised information to function effectively in different environments. A cleaning robot trained in Shenzhen may perform well in local offices but could fail in a Kuala Lumpur mall due to differences in spatial design and human activity, illustrating the need for region-specific data. Since 2020, Beijing has formally classified data as an economic asset on par with land, labour and capital. It has since supported the creation of domestic data exchanges where firms trade insights on consumer behaviour, industrial trends and more. SZDEX, a state-owned enterprise under the Shenzhen municipal government, operates one such platform. The collaboration with Zetrix AI is among the first attempts to extend this model beyond China's borders. No details were provided about the specific data types to be shared or the regulatory frameworks that will govern the exchange. The initiative focuses on creating secure mechanisms for cross-border data flows while generating commercial value. The next phase will involve technical and legal assessments to determine feasibility. What happens next depends on the outcome of these joint evaluations.
The Shenzhen Data Exchange is seeking international expansion while operating under a Chinese regulatory framework that restricts foreign access to domestic data. This creates a one-way dynamic where Chinese firms may gain insights from Southeast Asian markets without reciprocal transparency. The May 22 agreement with Zetrix AI does not specify safeguards for how Malaysian data will be used or protected. Without mutual data access, the partnership risks reinforcing asymmetric digital dependencies.
Editorial note: AI-assisted opinion, not established fact. Full disclaimer →
