Goldman CEO sees election volatility around AI data centers

Goldman Sachs CEO David Solomon anticipates potential political volatility surrounding data centers as the U.S. midterm elections approach. He noted that while there is significant discussion leading up to an election, the actions of states and localities post-election can create uncertainty.

Data centers, crucial for powering AI models, are facing increasing opposition from communities concerned about their substantial electricity demand, potential utility bill increases, and broader environmental impacts. Solomon also pointed out that the growing financing needs for AI projects, coupled with pressure on interest rates, could affect the returns and feasibility of certain initiatives.

According to a study by Stijn Van Nieuwerburgh, a finance and real estate professor at Columbia Business School, the expansion of AI infrastructure is on track to consume approximately 3.6% of the U.S. gross domestic product annually through 2032, representing over $10 trillion. This buildout involves increasingly complex financial arrangements that could pose systemic risks.

Solomon expressed confidence in Goldman Sachs' position to provide a full range of services for this growing sector, citing "enormous amount of activity." The firm is a financial partner with Nvidia, which announced in August a collaboration with six major financial institutions to launch compute platforms aimed at raising over $500 billion in third-party capital for AI infrastructure.