Google Scales Back AI Power Usage to Counter Grid Strain

Google is scaling back its power usage for AI data centers as the US electrical grid becomes strained by a surge in demand, primarily driven by the growth of artificial intelligence. The company has announced new agreements with utility partners to implement "demand response" capabilities, a strategy that allows it to shift or reduce its power consumption during periods of grid stress.

The International Energy Agency (IEA) estimates that nearly half of the expected increase in US electricity demand through 2030 will come from data centers, largely due to AI. This has put pressure on aging grid infrastructure and created challenges for utilities trying to keep up.

To address this, Google is for the first time applying demand response to its energy-intensive machine learning (ML) workloads. This builds on previous demand response efforts where the company would shift non-urgent tasks, like processing a YouTube video, to other times or locations. The company has announced new agreements with Indiana Michigan Power (I&M) and the Tennessee Valley Authority (TVA) to integrate these capabilities.

"As we add new large loads to our system, it is critical that we partner with our customers to effectively manage the generation and transmission resources necessary to serve them," said Steve Baker, President and COO, I&M. "Google's ability to leverage load flexibility as part of the strategy to serve its load will be a highly valuable tool to meet its future energy needs."

Google’s demand response program is designed to deliver several benefits. It allows large electricity loads like data centers to be interconnected more quickly, helps reduce the need for new transmission lines and power plants, and enables grid operators to manage power systems more efficiently. According to Google, this approach can help manage AI-driven growth even in areas where power generation and transmission are constrained.

Google acknowledges that data center demand flexibility is in the early stages and has limitations, as high levels of reliability are critical for services like Search and Maps, and for Cloud customers in essential industries. The company stated that a portfolio of solutions will be required to manage data center load growth, including investments in new generation and transmission. However, it believes demand response will play an important role in supporting a clean, reliable, and affordable energy system.

UN Urges Renewables Use for AI, Data Center Energy Demands

The UN is urging a rapid transition to renewable energy to meet the growing demands of AI and data centers, which are consuming a rapidly increasing share of global electricity. According to UN Secretary-General António Guterres, this shift is critical for both climate action and economic security. Citing a recent UN report, Guterres noted that while clean energy investments are outpacing fossil fuels, the transition is not happening fast enough or fairly enough. He called on governments to commit to aggressive new climate plans and urged major tech companies to power their data centers with 100% renewable energy by 2030, while also investing in modern grid infrastructure to handle the new energy load. He also highlighted AI's potential to improve the energy sector's efficiency and grid management