能源安全迎来“AI变量”，报告呼吁纳入“算力-电力弹性系数”

Core Insights - The report emphasizes that "reshaping energy demand under AI expansion" is a core variable for medium to long-term energy security governance, urging the inclusion of "computing power-electricity elasticity coefficient" in national energy strategy constraints [1] Group 1: Energy Security Challenges - China is facing internal challenges such as resource endowment constraints (rich in coal but poor in oil and gas), regional supply-demand mismatches, and an underdeveloped electricity market, alongside external shocks like fragmented energy trade and intensified geopolitical competition [2] - The report identifies that AI-related power consumption could increase by 33%-50% with each doubling of computing power, potentially leading to a "soft decoupling" of energy consumption from GDP by 2035, which would alter traditional peak paths [2] Group 2: Energy Governance Framework - The report proposes a new framework for energy governance, suggesting the integration of AI power demand into national energy planning and the implementation of a unified electricity market that connects long-term, spot, and ancillary service chains [3] - It advocates for a shift from "subsidy-driven" to "value-driven" approaches, including trialing carbon taxes to provide stable funding for renewable energy subsidies and constructing a diversified green investment system [3] Group 3: Future Energy Landscape - By 2035, the report predicts that China's energy security will exhibit four characteristics: safety and stability, green low-carbon, intelligent efficiency, and open collaboration, with non-fossil energy expected to account for 30% of the energy mix and a 50% reduction in energy consumption per unit of GDP compared to 2020 [3] - The report highlights the need for continuous monitoring and model iteration around the "computing power-electricity elasticity coefficient" to support China's dual carbon goals and modernization efforts [4]