The United States aims to invest over $10 trillion in AI infrastructure by 2032, representing approximately 3.22% of the country's annual GDP. However, the sector faces significant hurdles, including rising construction costs, project delays, and local opposition to new data centers.
According to RegLayer, at least 47 major projects were blocked, canceled, or delayed this year. Analysts from Bernstein estimate that only about 35% of planned data center projects in the US are considered viable. Furthermore, over 1,000 public groups have protested against data center construction in 2026.
Financial constraints are impacting major players. Nebius has secured funding for only 20% of its 5 GW capacity goal, which requires $250 billion. Meanwhile, SB Energy, backed by OpenAI and Nvidia, requires $174 billion in capital to fulfill contracts and has postponed its IPO, which was expected to value the company at $50 billion.
Market volatility and high interest rates are affecting companies like CoreWeave, whose shares recently dropped by over 25%. Concerns regarding the sustainability of the AI boom have led firms like OpenAI to delay public offerings. Additionally, many projects now require independent power generation due to grid limitations, further complicating development.
