Lina Hope
The relentless surge in artificial intelligence development, a transformative force reshaping industries worldwide, is encountering a fundamental physical constraint: the global electricity grid. As technology giants forge increasingly sophisticated AI models and deploy vast, energy-hungry computing infrastructures, the digital revolution is precipitating an unprecedented demand for physical energy. At the epicenter of this burgeoning challenge lies East Asia, the industrial powerhouse that underpins the global hardware supply chain.
Mainland China, home to the world’s second-largest data center market and the preeminent electronics-manufacturing hub, currently produces approximately 36% of global electronics output. This dominance extends to critical components for AI infrastructure. Taiwan stands as the undisputed leader in the production of the most advanced semiconductors, the very foundation of modern computing, and also serves as the primary manufacturing base for AI servers. Meanwhile, South Korea spearheads the global memory-chip industry, with giants like Samsung Electronics and SK hynix holding commanding positions in DRAM and NAND flash memory. Together, these three East Asian economies occupy strategically vital nodes across the data center, semiconductor, server manufacturing, and memory chip sectors, all of which are indispensable to the advancement and deployment of artificial intelligence.
However, this rapid expansion of computing power comes with a significant environmental toll. In 2024 alone, data centers globally consumed an estimated 415 terawatt-hours (TWh) of electricity, accounting for roughly 1.5% of total global electricity consumption. The International Energy Agency (IEA) reports that this figure has seen an annual growth rate of approximately 12% over the preceding five years. The implications for future demand are staggering. Greenpeace East Asia projects that the global electricity demand specifically for AI chip manufacturing could surge by as much as 170-fold between 2023 and 2030. Given the concentration of logic and memory chip production for AI hardware in East Asia, a substantial portion of this intensified manufacturing burden is poised to fall upon Taiwan, South Korea, and Japan.
Recognizing the urgency to address this challenge and chart a more sustainable path forward, Greenpeace East Asia convened a global webinar on August 6. The virtual forum brought together leading experts in energy policy and climate technology from the United States, Taiwan, and South Korea. The primary objective was to pivot the discourse from the necessity of decarbonizing AI to the practical strategies for building a clean and resilient AI supply chain.
Dr. Junyan Liu, deputy program director at Greenpeace East Asia, set the tone in her opening remarks, stating, "We would love to build shared momentum for a clearer and more sustainable AI and technology industry. We strongly believe that by partnering directly with industry leaders, we can transform complex environmental hurdles into viable, high-impact business models for the long haul." This sentiment underscores a proactive approach, seeking collaboration rather than confrontation to achieve a sustainable future for the digital economy.
Scaling Clean Technologies: Lessons from China’s Industrial Policy for the AI Era
Bridging the burgeoning electricity demand gap for AI necessitates a deep understanding of how clean technologies can be effectively scaled at an industrial level. China’s remarkable ascent in the sectors of solar energy, electric vehicles (EVs), battery manufacturing, and grid infrastructure offers a compelling blueprint for rapidly deploying clean technologies in the face of escalating demand.
Dr. Kyle Chan, a Fellow at the Brookings Institution’s John L. Thornton China Center, elaborated on China’s strategy, highlighting how the nation achieved end-to-end dominance across the value chain, from raw material extraction to finished component production. This was accomplished through a deliberate process of absorbing foreign technology, coupled with rapid adaptation and the co-development of robust local supply chains. Proactive local governments played a crucial role in accelerating this transition by providing essential infrastructure, land, and vital policy support.
"China’s cleantech policy wasn’t a single linear path toward becoming a powerhouse," Dr. Chan explained. "It involved constant pivots, strategic shifts, and adaptation along the way. China made a broad bet across multiple technologies, and it paid off." This diversified approach, rather than pinning hopes on a single technological winner, allowed Chinese manufacturers to simultaneously support various market segments. This strategy, combined with economies of scale and supply chain flexibility, proved instrumental in driving down costs. Furthermore, China’s established heavy industrial base created synergistic opportunities with its burgeoning clean-energy sectors. For instance, its shipbuilding capabilities have been leveraged to construct offshore-wind installation vessels, demonstrating an integrated approach to industrial development.
More recently, China has begun applying similar principles to the development of its data center infrastructure. Under the ambitious "East-West Computing Resources Transmission Project" (EWCRT Project), national policies are actively encouraging companies to relocate energy-intensive and less latency-sensitive computing workloads to the resource-rich western provinces. These regions offer abundant land and significant wind and solar energy potential. The processed data and resulting services can then be efficiently delivered to users in the densely populated coastal cities via a sophisticated national computing network. This strategic redistribution aims to optimize energy utilization and leverage renewable resources, mitigating the concentrated environmental impact on more developed eastern regions. This integrated approach, driven by industrial policy and strategic resource allocation, could offer invaluable lessons for other nations striving to power the AI era more sustainably.
The Economics of AI Power: Why Renewables Hold the Key in South Korea
In South Korea, the surge in demand from AI infrastructure presents a significant catalyst for accelerating the nation’s transition to clean energy, according to Jongkyu Kim, founder and chief executive officer of 60Hz and chair of the Korea Climate Tech Association. South Korea has recently announced substantial investments in three mega-projects, collectively valued at approximately 1,500 trillion won (roughly US$1 trillion). This comprehensive package includes the construction of four new semiconductor fabrication plants by Samsung Electronics and SK hynix, regional robotics initiatives, and an ambitious target of establishing 18.4 gigawatts (GW) of AI data center capacity by 2035. This projected capacity is equivalent to roughly 18-19% of South Korea’s recent peak electricity demand.
"AI has become an electricity story," Kim stated, echoing sentiments from industry leaders like Sam Altman, who suggested last year that "the cost of intelligence should eventually converge to near the cost of electricity." Kim elaborated, "If true, AI competition becomes a race for the cheapest electricity, and future-proof sources will win."
Kim’s analysis posits that renewables offer the most compelling economic case when compared to traditional sources like fossil gas or nuclear power. Data from Lazard’s 2026 Levelized Cost of Energy (LCOE+) Report illustrates this point clearly: on a dollar-per-megawatt-hour basis, unsubsidized renewable energy consistently emerges as the most cost-competitive form of new-build generation. This economic advantage is particularly significant for the long-term operational costs of energy-intensive AI infrastructure.
"AI is not a burden on the energy transition; it is a catalyst," Kim asserted. "Korea has anchored its AI build-out to renewables by location, law, and policy. Industry leaders have signed on to RE100, and the underlying economics close the loop. If the cost of AI is the cost of power, zero-marginal-cost renewables will win." This perspective highlights the synergistic relationship between AI growth and the renewable energy transition, suggesting that the pursuit of cheaper, cleaner energy sources will inherently drive the adoption of AI technologies.
Overcoming Bottlenecks in Taiwan’s Critical Chip Industry
As a preeminent global producer of advanced semiconductors and AI servers, Taiwan plays an indispensable role in powering the world’s computing infrastructure. Allissa Tai, RE100 representative in Taiwan at the Chung-Hua Institution for Economic Research (CIER), provided critical insights into the structural obstacles and emerging opportunities that manufacturers face in securing reliable renewable electricity.
Taiwan’s Ministry of Economic Affairs forecasts an average annual electricity demand growth of 2.5% from 2026 to 2035, a significant portion of which is attributable to the expansion of semiconductor fabrication plants, AI data centers, and increased cooling requirements. CIER estimates that AI and semiconductor-related projects alone could necessitate an additional 4.32 GW of power capacity by 2030.
In response to this projected surge, Taiwan’s Ministry of Economic Affairs plans to augment its energy supply by adding approximately 26 GW of new natural-gas-fired capacity by 2035. However, this strategy raises concerns given Taiwan’s heavy reliance on imported energy, with over 94% of its energy needs met through imports in 2024. An increased dependence on imported fossil fuels would expose the island to greater price volatility in international markets, potential shipping disruptions, and broader geopolitical risks, thereby jeopardizing its energy security.
While major technology corporations like Google and Apple are actively pursuing 100% renewable electricity for their operations and supply chains, the corporate procurement of RE100 members currently accounts for only about 5% of Taiwan’s total renewable generation, according to data compiled by CIER and the Climate Group’s 2024-2025 Taiwan Renewable Electricity Market Briefing. Several key structural obstacles hinder a more widespread adoption of renewables:
- Limited Supply of Renewable Electricity: The existing renewable energy infrastructure and the rate of its expansion are insufficient to meet the rapidly growing demand from energy-intensive industries.
- Inadequate Corporate Power Purchase Agreements (PPAs): The market for long-term, stable PPAs that offer price certainty for renewable energy is still developing, making it challenging for companies to secure sufficient green electricity at competitive rates.
- Regulatory and Permitting Hurdles: Complex regulatory frameworks and lengthy permitting processes can delay the development and deployment of new renewable energy projects, further constraining supply.
- Grid Infrastructure Limitations: The existing grid infrastructure may require significant upgrades to accommodate the intermittency of renewable sources and the concentrated demand from large industrial users.
"To power the AI boom cleanly, we need energy infrastructure built at scale," Tai emphasized. "That requires policy stability, social consensus, and cross-party alignment. Transitioning to renewables is not just about climate; it is the bedrock of Taiwan’s economic security and global competitiveness." This statement underscores the critical intersection of environmental responsibility, economic prosperity, and national security in Taiwan’s energy future.
An Action Plan for a Clean AI Ecosystem
The convergence of escalating AI development and the imperative to align with global climate goals necessitates a clear and actionable strategy. The panelists at the Greenpeace East Asia webinar identified three key areas for targeted reform to foster a clean AI ecosystem:
1. Accelerating Renewable Energy Deployment and Infrastructure Investment
There is a critical need to significantly scale up the deployment of renewable energy sources, including solar, wind, and other sustainable technologies. This requires substantial investment in grid modernization and expansion to ensure the reliable integration of these sources. Governments should prioritize policy frameworks that incentivize private sector investment in renewable energy projects, streamline permitting processes, and facilitate the development of long-term power purchase agreements (PPAs) that offer price stability and predictability for large energy consumers. Examples from countries like Germany and Denmark, which have successfully integrated high percentages of renewables into their grids, can offer valuable lessons.
2. Enhancing Energy Efficiency and Demand-Side Management
Beyond increasing renewable supply, a parallel focus on energy efficiency across the entire AI value chain is paramount. This includes optimizing data center design and operations for lower energy consumption, developing more energy-efficient AI algorithms and hardware, and implementing intelligent demand-side management strategies. Technologies such as advanced cooling systems, server virtualization, and workload optimization can significantly reduce the energy footprint of AI operations. Furthermore, promoting the use of less energy-intensive computing models and encouraging responsible AI development practices will be crucial.
3. Strengthening International Cooperation and Supply Chain Transparency
Given the global nature of the AI supply chain, robust international cooperation is essential. This involves sharing best practices, harmonizing standards for energy efficiency and renewable energy procurement, and fostering collaborative research and development in clean energy technologies for computing. Increased transparency throughout the AI supply chain, from chip manufacturing to data center operations, will enable stakeholders to identify and address environmental hotspots effectively. Initiatives like the RE100 campaign, which encourages companies to commit to 100% renewable electricity, can be expanded and strengthened through greater international coordination and support from governments and financial institutions.
By implementing these interconnected strategies, the global community can work towards building an AI future that is not only technologically advanced but also environmentally sustainable and economically resilient, ensuring that the digital revolution contributes positively to both human progress and planetary well-being.
Raul Delapena Setiawan
Nila Kartika Wati
