BIS Paper Says AI Power Demand May Affect Inflation, Output and Financial Stability

Artificial intelligence’s rising electricity demand may affect monetary-policy variables, including productivity, inflation and potential output, while large-scale AI and energy infrastructure investment may have financial-stability implications, a Bank for International Settlements paper published Oct. 8 said.

The paper, titled “Artificial intelligence and climate change: balancing innovation and sustainability,” examines how computing demand interacts with energy systems and climate conditions. Leonardo Gambacorta and Salvatore Polizzi wrote the study.

The paper says the AI-climate nexus can influence potential output and inflation dynamics. Higher productivity could change the economy’s capacity to produce goods and services, while increased power demand and infrastructure spending could affect prices and investment conditions.

It presents two possible paths for adoption: an “AI copilot” scenario involving gradual integration and a more transformative scenario involving artificial general intelligence. The paper does not quantify the probability or potential size of a financial-stability event.

Global data center electricity consumption was about 415 terawatt-hours in 2024, equal to roughly 1.5% of worldwide electricity use. In a base case, consumption could reach about 945 terawatt-hours in 2030, just under 3% of global electricity consumption.

Electricity use from accelerated servers, driven mainly by AI adoption, is projected to grow 30% annually from 2024 through 2030, compared with 9% annual growth for conventional servers. The United States is projected to add about 240 terawatt-hours of data center consumption over the same period, a 130% increase from 2024.

The paper also identifies potential benefits from AI, including more efficient energy use, improved resource allocation, faster climate innovation and stronger climate forecasting, adaptation and risk management.

For crypto markets, the relevant channels are interest-rate expectations, energy costs and broader risk appetite. Higher power demand could add to inflation pressure in some regions, while weaker AI infrastructure investment could affect financial conditions. Those are potential policy transmission channels, not evidence of a current crypto-market shock.