Page 33 - The Annual AI Governance Report 2025 Steering the Future of AI

The Annual AI Governance Report 2025: Steering the Future of AI

sources may be difficult. We may find increased short-term reliance on fossil-fuel based energy
sources.
120
Renewable Energy Transition: Artificial Intelligence could potentially support the achievement
of SDG targets by enhancing renewable energy systems through optimization, forecasting,
automation, and smart communication. AI could help all five targets of Affordable and Clean
Energy by improving grid efficiency, reducing costs, and integrating intermittent sources like
solar and wind. AI contributes to climate action by optimizing emissions reductions and energy
planning. In the economic domain, it enhances productivity and innovation by supporting green
jobs, efficient industry, and waste-to-energy systems.
121
AI-enabled smart grids use machine learning and real-time data analytics to optimize the
generation, distribution, and consumption of electricity, which is crucial for integrating variable
renewable energy sources like wind and solar. These smart grids can forecast energy demand and
RE supply with high accuracy, enabling better scheduling and load balancing. AI helps manage
distributed energy resources (e.g., rooftop solar, electric vehicles, batteries) by automating
decisions about when to store, release, or redirect energy. This reduces reliance on fossil fuel
backup systems, cuts emissions, and enhances grid stability and resilience. Additionally, AI can
detect faults, predict equipment failures, and respond to outages faster than traditional systems,
lowering costs and improving reliability.
122
Green AI is a paradigm focused on reducing the environmental impact of artificial intelligence
systems while maintaining performance. Green-in-AI, which involves designing energy-efficient
models and algorithms that minimize computation, and Green-by-AI, which uses AI to advance
sustainability in other sectors like energy, agriculture, and climate policy. Green AI emphasizes
123
strategies such as algorithm optimization, efficient hardware (like TPUs), and edge computing to
reduce emissions. Regulations like the EU’s AI Act now mandate reporting energy use for high-
risk AI systems; tools like CarbonTracker and CodeCarbon are used to measure emissions. 124

5.3 Hardware Innovation and Supply Chains

AI Chip Supply Chain: AI supply chain risk is highly centralized, where the production of
advanced AI chips depends heavily on a few key players. The supply chain resembles an "inverted
triangle," with one global supplier of EUV lithography machines at the narrow base, and another
which produces around 90% of AI chips, as the critical bottleneck. This concentration of power
makes the supply chain vulnerable to geopolitical tensions.
125 126

120 McKinsey & Company (2025), How data centers and the energy sector can sate AI’s hunger for power.
121 Hannan, M., Al-Shetwi, A. Q., Ker, P. J., Begum, R., Mansor, M., Rahman, S., Dong, Z., Tiong, S., Mahlia, T. I.,
& Muttaqi, K. (2021). Impact of renewable energy utilization and artificial intelligence in achieving sustainable
development goals. Energy Reports, 7, 5359–5373.
122 Ahmad, T., Madonski, R., Zhang, D., Huang, C., & Mujeeb, A. (2022). Data-driven probabilistic machine
learning in sustainable smart energy/smart energy systems: Key developments, challenges, and future
research opportunities in the context of smart grid paradigm. Renewable and Sustainable Energy Reviews,
160, 112128.
123 Bolón-Canedo, V., Morán-Fernández, L., Cancela, B., & Alonso-Betanzos, A. (2024). A review of green artificial
intelligence: Towards a more sustainable future. Neurocomputing, 599, 128096.
124 Thakur, D., Guzzo, A., Fortino, G., & Piccialli, F. (2025). Green Federated Learning: A new era of green aware
AI. ACM Computing Surveys.
125 Gopal, S., Staufer-Steinnocher, P., Xu, Y., & Pitts, J. (2022). Semiconductor Supply Chain: A 360-Degree
view of supply chain risk and network resilience based on GIS and AI. In Springer series in supply chain
management (pp. 303–313)
126 Mison, A., Davies, G., & Ward, R. (2024). Cross-disciplinary AI supply chain risk assessment. Reading:
Academic Conferences International Limited.

24