The rise of artificial intelligence (artificial intelligence) is squeezing resources, raising concerns about the need for labor, copper, water, and other basic necessities to build and operate data centers across the globe. However, the financial commitment to these projects, which are already in planning stages, is high. Nvidia boss Jensen Huang estimates it costs $60 billion to $80 billion to construct a 1 GW compound housing thousands of server racks, compared to a $36 billion per GW figure from Bernstein analysts, which would translate into $4 trillion to pay for all the currently slated data centers. Without government funding, private investors will need to contribute to the technology's expansion. The cost of these expansion is also considered higher than most estimates, with some suggesting it could reach $6.6 trillion. The surplus assumes that hyperscalers like Microsoft will devote all of their operating cash flow to data centers for the next five years, but interest payments on associated debt will add significant amounts. Additional lending capacity is also expected to be provided by high-yield bonds and leveraged loans, and Morgan Stanley analysts predict that about $50 billion annually from asset-backed and commercial mortgage-backed securities related to building AI-related facilities will provide another $150 billion through 2030.