The EU AI Act's August 2025 deadline applies specifically to providers of general-purpose AI (GPAI) models — systems that can perform a wide range of tasks across different domains. This includes foundation models used in enterprise software, advisory tools, and automated decision systems.
For companies operating in both Europe and Asia, the compliance challenge is asymmetric: the EU mandates transparency, risk classification, and human oversight, while China and Southeast Asia operate under different — and rapidly evolving — AI governance frameworks.
The practical implication: cross-border organizations cannot apply a single compliance architecture. They need jurisdiction-specific legal and technical structures, with a clear data residency and model deployment strategy for each market.
PREXKON works with Dr. Kang Li's team to map this dual-compliance landscape for our clients.