AI financing and energy financing are accelerating their integration, with tech companies beginning to proactively invest in power generation assets or adopt “off-grid power” solutions. At the same time, labor shortages, water resource pressures, and policy opposition in multiple regions are further exacerbating the structural tightness in computing power supply.
As tech giants including Microsoft (MSFT), Google (GOOG), Meta (META), Amazon (AMZN), and Oracle (ORCL) continue to raise their capital expenditure budgets, the scale of artificial intelligence infrastructure construction is heading toward historic highs. Market attention has largely focused on data center expansion, graphics processing unit (GPU) procurement, and enormous financing demands. However, Morgan Stanley’s latest research points out that electricity supply is becoming a key bottleneck for AI expansion, and the boundary between AI financing and energy financing is thus rapidly blurring.
A few weeks ago, Morgan Stanley explored the topic of AI infrastructure financing, analyzing how tech companies can raise funds for massive expenditures through equity and debt markets. However, after in-depth discussions with clients and industry chain companies, the firm has become increasingly convinced that the financing issue can no longer be understood solely from within the technology sector. The bottlenecks for AI infrastructure are extending upstream: while data centers are the ultimate vehicles for delivering computing power, the electricity, transmission networks, transformers, power generation equipment, and even labor and water resources that support their operation are becoming equally important constraints. AI infrastructure financing is shifting from pure data center financing to an integrated “electric power + computing power” financing model.
Convergence of AI Financing and Energy Financing, with Labor, Water Resources, and Policy Also Posing Challenges
As electricity constraints intensify, the boundary between AI and energy infrastructure is rapidly dissolving. In the past, energy projects were mostly invested in by utility companies or independent power producers, while tech companies focused on servers and chips. Now, to secure stable power supply, a growing number of AI companies are proactively participating in energy asset investments, including signing long-term power purchase agreements, directly investing in power generation projects, and even acquiring related assets.
Moreover, broader structural constraints cannot be overlooked. On the labor front, the United States is projected to face a shortage of approximately 300,000 electricians over the next decade, and more than one-fifth of the existing electricians are already aged 55 or above.
Amid multiple constraints, Morgan Stanley believes the risk of structural imbalance between computing power supply and demand is on the rise. If AI demand continues to grow rapidly while electricity, equipment, labor, water resources, and policy approvals constrain supply expansion, “scarcity” will become a prominent feature of the computing power market. Companies with access to large-scale, reliable computing power resources will gain stronger pricing power. The report refers to such companies that possess scarce, stable, and deliverable computing capacity as “computing power suppliers,” whose core advantage lies not merely in data centers or cloud services, but in their command over critical resources. Similar to power generation companies that hold key resources in the energy industry, their bargaining power within the industry chain is expected to increase significantly.
