How much electricity will South Korea need in five minutes? Power supply can change with heat waves or cloud cover over solar power plants. With the increase in data centers and electric vehicles, the calculations have become more complex. Kim Sung-jin, chairman of the Korea Power Exchange, is building an 'AI control tower' for predicting supply and demand using AI and data.
Electricity is different from regular commodities. The amount used and produced nationwide must be balanced at every moment. A shortage increases the risk of blackouts, while excess generation raises costs.
The Korea Power Exchange manages this massive balance. It forecasts national electricity demand, plans generation, and operates the power system and market. Although it may not be visible to the public, it serves as the control tower for South Korea's power system.
AI Reads 'Electricity in 5 Minutes'
The core task of the Korea Power Exchange is forecasting electricity demand. It must predict how much electricity will be needed at 10 a.m., 3 p.m., and 8 p.m. tomorrow to prepare generation. Numerous variables, including temperature, humidity, day of the week, time, industrial activity, and holidays, influence electricity usage.
The exchange utilizes vast data to forecast daily and weekly electricity demand and is advancing AI-based predictions as a key strategic initiative. It also produces five-minute interval demand forecasts.
We are entering an era where data and AI can predict how much electricity South Korea will need in five minutes.
AI's importance is growing as solar and wind energy increase.
In the past, adjusting generation to match demand was centered around large power plants like nuclear, coal, and gas. Now, cloud cover can reduce solar output, and weak winds can decrease wind generation.
It is essential to predict not only how much electricity will be consumed but also how much will be produced.
With the rise of energy storage systems (ESS), electric vehicles, and distributed energy resources, the power grid is becoming a vast network connecting numerous suppliers and consumers in real-time. This is why AI is necessary.
AI Consumes Electricity and Directs It
A new variable shaking up the power market is AI itself. The spread of generative AI has led to an increase in large data centers. Operating numerous GPUs is required to train and service AI. When large data centers are established, they create significant new electricity demand in the area.
Ironically, while AI consumes vast amounts of electricity, it is also needed to ensure a stable supply of that electricity.
In the future, the Korea Power Exchange will need to consider nuclear, gas, solar, wind, ESS, electric vehicles, and AI data centers simultaneously.
The complexity of the power system has grown too intricate to rely solely on human experience and existing statistics. The value of AI in accurately predicting supply and demand and supporting optimal choices is becoming increasingly significant.
Kim Sung-jin's 'Energy Transition'
Since taking office in May, Kim Sung-jin has identified building a stable power supply system, innovating the power market, and establishing a distributed power system as key tasks.
In particular, distributed energy is a crucial variable that will change the future power system. The traditional method of sending electricity produced at large power plants through long-distance transmission networks to metropolitan and industrial areas is shifting towards a system where electricity is produced and consumed locally.
The relationship between electricity and industrial location is also changing.
As facilities that consume vast amounts of electricity, such as semiconductor factories and AI data centers, continue to increase, there are limits to building transmission networks everywhere they are needed. A new approach is required to guide industries to areas with abundant electricity.
Regional electricity prices, distributed energy, and industrial locations are becoming interconnected.
AI and data can be utilized to analyze how much electricity is needed in each region, how demand will grow, and what burdens will arise on the transmission network.
The role of the Korea Power Exchange may also expand from simple power supply management to supporting the efficient allocation of energy and industry through a data platform.
Korea Power Exchange's Transition to 'KPX AI'
The Korea Power Exchange itself has begun its own transformation. In response to the energy transition and AI innovation in 2026, it has restructured its organization and enhanced its AX data capabilities. It is also pursuing an information strategy plan for the introduction of 'KPX AI.' The core asset is data.
Vast amounts of data are accumulated during the processes of generation, electricity demand, market prices, and system operations. By making this data usable for AI and linking predictions and analyses to actual decision-making, the accuracy and efficiency of power system operations can be improved.
AI-based electricity demand forecasting is just the starting point. There is significant potential for expanding its application to renewable energy generation forecasting, power market analysis, and distributed resource management.
The essence of the AX that the Korea Power Exchange should aim for lies here. It is not about creating many AI systems but using AI to predict South Korea's electricity more accurately and operate it more stably.
The 'AI Control Tower' of South Korea's Electricity
As the AI era takes off, the role of South Korea's power industry is becoming clearer. While power companies like Korea Hydro & Nuclear Power generate electricity and Korea Electric Power transmits and distributes it, the Korea Power Exchange matches supply and demand nationwide and directs the entire system. It serves as the 'brain' of South Korea's electricity.
Now, AI is entering this brain.
With the increase of solar and wind energy, the spread of ESS and electric vehicles, and the emergence of large new electricity consumers like AI data centers, the complexity of the power system is growing.
The success of the energy transition that Kim Sung-jin is promoting depends on how stably and efficiently this complex system is managed.
AI predicts electricity demand and analyzes renewable energy generation. It reads vast amounts of power data to assist in the operation of the power system and market.
We are moving beyond an era where AI consumes electricity to one where AI directs South Korea's electricity. The Korea Power Exchange, the control tower of South Korea's electricity, is now evolving into an 'AI control tower.'
:Kim Sung-jin, Chairman of the Korea Power Exchange:
Kim Sung-jin is an expert in industrial and energy policy. He has served as the spokesperson for the Ministry of Trade, Industry and Energy, director of the Trade Policy Bureau, industrial and trade secretary at the Presidential Office, and director of Gwangju Technopark. He became chairman of the Korea Power Exchange in May 2026, promoting energy transition centered on stable power supply, power market innovation, and the establishment of a distributed power system.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.