Transition empowerment

Climate change is one of the greatest challenges for humanity because of its potential ecological consequences, environmental, sociopolitical and socioeconomic areas.

The main factor contributing to current climate change is greenhouse gas emissions caused by human activities, which have pushed the global surface temperature 1.1 C above the 1850-1900 level in the 2011-20 period. The Paris Agreement set the long-term objective of limiting global warming to well below 2 C (ideally 1.5 C) above pre-industrial levels.

In September 2020, China announced its dual carbon goals of peaking carbon dioxide emissions before 2030 and achieving carbon neutrality before 2060. The three pillars for building a low-carbon and even net-zero carbon emissions society are clean energy (renewable energy, nuclear energy), energy storage and conversion (hydrogen energy, smart grids), and low-carbon industry and manufacturing (carbon capture, utilization and storage). China's domestic green and low-carbon transition, which focused on these three pillars under the 14th Five-Year Plan (2021-25), has made remarkable progress.

In 2024, the increase in the global average annual temperature exceeded 1.5 C above pre-industrial levels for the first time, according to the World Meteorological Organization data. Recent assessments also reveal that the global remaining carbon budgets for achieving the 1.5 C and 2 C targets are below 205 billion metric tons CO2 and 860 to 955 billion tons CO2, respectively. By 2030, the storage capacity may reach around 670 million tons CO2. Nevertheless, the current pace of decarbonization remains insufficient to meet the net-zero goal. The window to reach the net-zero emissions by the middle of the century is closing fast.

Artificial intelligence, which has developed rapidly with the continuous exploration, research and development of data, algorithms and hardware, can play a crucial role in accelerating the process of decarbonization. The Recommendations of the 20th Central Committee of the Communist Party of China for Formulating the 15th Five-Year Plan (2026-30) for the National Economic and Social Development promote the deep integration and broad application of "AI+" in the country's green and low-carbon strategies. This means that more efforts will be needed to apply big data, algorithms and computing power to promote the development of low-carbon industries, clean energy, smart manufacturing and green cities in the next five years.

Although AI itself (mainly data centers) accounts for 0.5 percent of total CO2 emissions by consuming 1.5 percent of global electricity and 0.08 percent of global freshwater (and the amount will double by 2030), it remains a potentially pivotal tool for decarbonization by overcoming barriers of complexity, speed and scale. But releasing AI's full potential requires consideration of the following actions.

First, investing in AI for decarbonization. Capital and resources should be directed toward technologies, companies and systems where AI can accelerate emissions reduction across energy, transportation, buildings and natural ecosystems. Recognizing that financing is essential to the energy transition, China leads the world in energy transition investment, accounting for two-thirds of the $2.1 trillion spent globally in 2024, according to a report by Bloomberg New Energy Finance. According to the International Data Corporation's Worldwide AI and Generative AI Spending Guide, China's total AI investment will surpass $100 billion by 2028, which will account for 7 percent of the energy transition investment.

Second, improving data quality and sharing. High-quality, structured and labeled data are the foundation for AI performance. Data sharing is essential for releasing the full potential of AI. Enabling responsible data sharing across institutions, industries and regions improves performance, reduces bias and enhances AI models. This may involve accuracy, consistency, cleanliness, diversity, privacy, ethics, documentation standardization, licensing and policy. Good examples are the Materials Project and the ImageNet, which provide datasets on the properties of all inorganic materials and labeled images. These datasets dramatically promote AI-driven materials discovery and computer vision. Similar ambitious datasets are also required for decarbonization.

Third, deeply integrating AI and industrial applications. AI can accelerate decarbonization by strengthening forecasting, maintenance and operational efficiency in clean energy, improving electrolyzer control, storage dispatch, load forecasting and grid stability, as well as energy storage and conversion, optimizing capture processes, designing transport and storage networks, and enabling real-time monitoring and verification in carbon capture, use and storage. Deep integration requires strong data infrastructure, digital twins, intelligent control and human oversight, enabling industries to transition to low-carbon systems.

Fourth, applying responsible AI for decarbonization. Decarbonization relies heavily on AI for forecasting renewable energy, optimizing hydrogen production, guiding grid operations and improving the performance of carbon-intensive industries. But errors, bias or misuse in these systems can lead to operational failures, safety risks or inefficient energy use that undermines climate goals. Implementing responsible AI ensures that AI-driven solutions are accurate, trustworthy, equitable and safe as they are deployed across critical energy and industrial systems. Implementing responsible AI for decarbonization concerns high-quality, secure, and representative data, explainable AI decisions, validation of stress testing, human oversight, robust cybersecurity and ethical guidelines.

Fifth, strengthening international cooperation. International cooperation faces significant obstacles because of export barriers, data restrictions, divergent carbon standards and geopolitical confrontation. China should simultaneously strengthen its independent innovation capability in technology, focus on pragmatic cooperation in low-sensitivity, high-consensus areas such as methane monitoring, open-source climate models and academic datasets, actively shape global rules and encourage multinational enterprise collaboration to bypass geopolitical barriers, and expand cooperation between Global South countries.

On the way to the net-zero goal, AI is not a panacea — but without it, decarbonizing is unlikely to be fast enough.

Chen Chong is an associate professor at the College of Artificial Intelligence at China University of Petroleum (Beijing). Rui Zhenhua is the dean of the College of Geophysics and the director of the International Innovation Research Institute for Carbon Capture, Utilization and Storage at China University of Petroleum (Beijing). The authors contributed this article to China Watch, a think tank powered by China Daily.

The views do not necessarily reflect those of China Daily.

Contact the editor at editor@chinawatch.cn.