Schneider Electric CEO: AI could cut energy consumption by 30%

Schneider Electric CEO Olivier Blum said that artificial intelligence (AI) has the potential to reduce energy consumption by up to 30% through automation of power systems.

Speaking at the India AI Impact Summit in New Delhi, Blum emphasized that managing demand, rather than supply, could be key to addressing global power issues.

He noted that global power demand from data centers is expected to nearly triple to 219 gigawatts by 2030, with AI workloads making up about 70% of that growth, according to McKinsey & Co.

Blum highlighted that current AI technology is only about 20% developed for energy management and described the next two years as critical for progress.

The expansion of renewable energy sources has increased the need for smarter grid management, and AI could play a significant role in making energy systems more flexible.

🔗 Source: Bloomberg

🧠 Food for thought

Implications, context, and why it matters.

AI’s energy savings are already being realized in specific sectors

The CEO’s 30% savings claim remains a broad target, yet several AI uses already deliver measurable cuts.
In commercial buildings, AI-powered building management systems can run heating and lighting based on real-time occupancy plus weather, trimming energy costs by up to 40% 1.
Factories apply predictive AI to tune steam and cooling systems, which can take up half of a plant’s energy use, lowering consumption by as much as 10% 2.
Data centers also use AI to curb their own load; a field test on a GPU (graphics processing unit) cluster reduced power use by about 25% for several hours at peak times by shifting workloads during peak grid demand 3.

AI is in a race against its own energy consumption

Efficiency efforts now track a growing energy squeeze, since global data center power demand is expected to nearly triple to 219 gigawatts by 2030, with AI workloads driving about 70% of the increase.
AI data centers could reach up to 12% of U.S. electricity use by 2028, adding stress to a grid where about 70% of equipment nears the end of its life cycle 4, 5.
That strain is shifting priorities from cutting energy alone to raising energy productivity, which has sparked metrics such as “tokens per watt per dollar” for model output 5.
Energy-heavy sites are taking a more active role with the grid by using on-site power systems to provide stability services and help utilities manage swings in renewable generation 6.