US Federal Reserve: AI Data Center Boom Resembles Pre-2008 Global Financial Crisis

US Federal Reserve officials are beginning to discuss the risks to financial stability posed by the artificial intelligence boom, according to Reuters. Federal Reserve Bank of San Francisco President Mary Daly stated that the pace and scale of investments in AI are a serious concern, particularly the rising level of leverage, and the Fed needs a 'dashboard' to monitor potential market failures. Federal Reserve Bank of Kansas City President Jeff Schmid echoed these concerns, questioning whether the circular commitments involving data centers and energy suppliers could lead to excessive risk-taking, and whether a spark could ignite a fire. Pessimists even compare the situation to the US housing market before the 2008 global financial crisis, with Schmid asking if the AI industry is becoming 'too big to fail'. However, Federal Reserve Bank of New York President John Williams disagrees, stating that this is not a bubble situation, as the borrowing for AI investments is managed by highly profitable companies. Torsten Slok, chief economist at Apollo, added that data center construction volumes remain less than half of the housing boom peak of 6.6% of GDP in 2005. About 850 data centers are being built worldwide, totaling around $7 trillion, with the US leading (228 projects) and China second (98 projects), according to Oxford Economics. In late 2025, US data center construction costs surpassed office building and even global oil industry investments. Turner Construction alone realized $9.4 billion in data center projects in 2025, more than five times its 2020 figure. The rapid buildout also strains power grids, causing higher electricity rates and local protests.