Introduction

Artificial intelligence (AI) has the potential to reshape energy systems globally, aiding the transition to renewable energy and causing an increase in electricity demand from data centres. In East Asia, data centre electricity consumption is forecast to more than double.

As the region produces approximately 90% of the world’s leading semiconductor chips, its role in the global AI supply chain is critical. East Asia also faces unique challenges, such as fragmented power grids with minimal cross-border interconnections and a strong reliance on imported fossil fuels.

Chapter 1. AI for energy

AI can optimise the deployment of variable renewable energy (VRE) by improving planning, siting, and permitting processes. For instance, wind farm design can be enhanced using deep learning to simulate atmospheric conditions and aerodynamic efficiency. In power grids, AI facilitates predictive maintenance, real-time grid operation, and dynamic line rating, which can increase transmission line capacity by 15-30%.

Furthermore, AI aids demand-side flexibility by forecasting consumption, coordinating virtual power plants, and managing electric vehicle charging. It also accelerates innovation in technologies such as battery chemistry, green hydrogen electrolysis, and carbon capture.

Chapter 2. Energy for AI

Data centres require substantial electricity for servers, cooling, and network equipment. The global electricity consumption of data centres is projected to more than double from 415 TWh in 2024 to 945 TWh by 2030. In China, data centre demand is expected to reach 280 TWh by 2030, constituting nearly 30% of global consumption.

AI operations, particularly model training, necessitate high-density server racks that require advanced liquid cooling systems. To meet these demands sustainably, major technology firms are increasingly securing renewable energy through power purchase agreements (PPAs), though matching energy supply with continuous data centre demand remains challenging. Alternative zero-carbon firm energy sources, such as nuclear power and advanced geothermal, are being actively explored.

Chapter 3. High-level policy recommendations

The report recommends accelerating the digitalisation of energy systems to enable rapid AI adoption. Investment frameworks should prioritise the rollout of sensors, data platforms, and control systems that allow large-scale data collection. Policy frameworks should be updated to establish clear guidelines on data privacy and cybersecurity without creating disproportionate compliance burdens.

Governments are advised to promote clean energy procurement by simplifying PPA regulations and facilitating investment in grid infrastructure. Additionally, policymakers should incentivise the siting of data centres in regions with available grid capacity and abundant renewable resources. Finally, improving data collection and transparency regarding data centre electricity use is crucial for effective power system planning.