From Paradox to Progress A Net Positive AI Energy Framework 2025

Behavioural and demand side –Unconscious consumption and rebound effects –Elastic demand from low marginal costs –Limited incentives for energy-efficient behaviour Regulatory and policy –Inconsistent regional policy frameworks –Slow policy adaptation to AI –Limited uptake of voluntary energy codes and standards Workforce and capacity –Skills gaps in AI-energy domains –Limited education on efficient AI –Uneven regional readiness Ecosystem fragmentation –Siloed innovation and interoperability gaps (technical, regulatory, etc.) –Misaligned incentives between stakeholders –Low trust hindering collaboration Addressing these challenges will require collaboration, investment and accountability across relevant market sectors. The opportunity: AI as an energy and climate asset Despite these challenges, AI holds immense potential to accelerate the clean energy transition while also driving competitiveness. Deployed responsibly, AI can enable many benefits, including: –Reducing data centre cooling energy use by 40%23 –Improving commercial building heating, ventilation and air conditioning (HVAC) energy efficiency by 15–40%24,25 –Shortening convoluted permitting processes26 –Optimizing grid operations, reducing losses and improving reliability27 –Enhancing forecasting for renewable energy integration28 –Streamlining logistics and industrial processes to cut emissions29 The opportunity is clear: if deployed with intention, AI can deliver net-positive energy and climate outcomes, where the benefits outweigh its energy consumption. Why a new framework is needed The current trajectory of AI development is largely growth-focused, emphasizing scale, speed and capability.30 This approach is no longer sufficient.31 According to industry stakeholders, ecosystem actors must instead shift to an impact-first paradigm that prioritizes measurable outcomes over raw performance. This framework does not call for constraint, but rather for strategic alignment, ensuring AI’s rapid growth advances innovation, supports sustainability and reinforces long-term resilience. Who is this framework for? Achieving net-positive AI energy demands collective action. This framework helps stakeholders ensure AI’s energy impact becomes a strategic advantage, not a liability. Those with mature AI systems are applying them to sustainability outcomes at multiple times the rate of others and with a stronger emphasis on long- term value creation.32 Building this level of capability across industries and regions is crucial to achieving net-positive energy outcomes at scale. The opportunity is clear: if deployed with intention, AI can deliver net- positive energy and climate outcomes, where the benefits outweigh its energy consumption. 9 From Paradox to Progress: A Net-Positive AI Energy Framework