By Arun Rai
As artificial intelligence(AI) powers a new wave of economic transformation, the physical infrastructure behind it—data centers—is colliding with a geopolitical reality check. In fast-growing tech hubs like Atlanta, where demand for AI compute is skyrocketing, developers and investors are running into bottlenecks not just in materials, but also in the global rules of engagement.
Tariffs, export bans, and investment restrictions are no longer background noise; they are frontline forces reshaping how and where digital infrastructure gets built. Data centers are no longer just part of the digital economy anymore; they are the physical backbone of AI infrastructure. As demand for model training and inference grows, so does the pressure to scale compute capacity faster, cleaner and closer to end users.
But the story is not just about constraint. For states like Georgia, these pressures create a rare opportunity to lead the next wave of AI growth by building resilience into the very backbone of the digital economy.
Strategic Arc: How AI Infrastructure Is Being Disrupted and Rebuilt
A once-optimized global supply chain is breaking down. U.S. tariffs, reaching 145 percent on some tech imports—combined with retaliatory restrictions from Beijing, have destabilized the sourcing of critical components. Export controls on Nvidia’s H100 chips and other advanced semiconductors are further squeezing availability. What was once a stable, cost-driven pipeline is now fraught with geopolitical volatility.
Atlanta is experiencing a surge in data center demand. In 2024, it surpassed Northern Virginia in net data center leasing—becoming the most active U.S. market for space brought online. Amazon Web Services has announced $11 billion in infrastructure investment across Georgia, and Microsoft is expanding its footprint in the region. But growth is now colliding with constraints: component delays, sourcing risk and capital friction are beginning to shape what gets built—and how fast.
Developers and investors are already pivoting. Supply chains are shifting away from China, with new assembly operations emerging in Mexico, Vietnam, and India. Investors are redesigning capital stacks to meet expanded scrutiny by the Treasury Department’s Committee on Foreign Investment in the United States . In today’s environment, flexibility isn’t just operational; it is strategic.
The deeper question is whether Georgia can maintain its data center leasing lead when global politics, not just market forces, shape what gets built. Tax incentives and power rates still matter, but agility in navigating policy and supply constraints will determine the next wave of winners.
How the Pressures Hit the Ground
With tariffs, hardware costs have risen by 7-12 percent on large projects—adding as much as $30 million in additional capital cost for a typical 36-megawatt data center in metro Atlanta. AI scalability also depends on data centers securing advanced hardware—especially high-performance chips and memory components. These chips are fabricated and assembled abroad, primarily by Taiwan-based TSMC (Taiwan Semiconductor Manufacturing Company), making them vulnerable to global supply bottlenecks and tariff pressures.
Atlanta remains well positioned: land availability, power infrastructure and access to fiber remain strong. The Port of Savannah strengthens logistics. But the region’s reliance on imported components and Chinese-origin equipment creates fragility that is hard to ignore. Strategic shifts, such as building transmission capacity or attracting alternative suppliers, will be essential.
What’s at Stake and Where Georgia Can Lead
This is not just a supply chain story. It’s a realignment of where, how and by whom AI infrastructure gets built. The regions that can rapidly adapt by localizing component sourcing, modernizing power infrastructure, and attracting vetted capital stand to gain the most in the next phase of the AI economy.
Georgia’s position is more than exposed: it’s leveraged. The state’s deep logistics backbone, led by the Port of Savannah, gives it a head start in rerouting global supply chains. As manufacturers shift server assembly to Mexico and Vietnam, Atlanta’s proximity and freight access position it as a natural redistribution hub for reconfigured tech supply lines.
On the capital front, the scrutiny facing foreign-backed projects puts a premium on trusted, compliant financing. Georgia could attract even more hyperscale investment by streamlining approvals for allied investors and demonstrating transparency in grid planning and infrastructure alignment.
With growing interest in domestic cooling, power systems, and grid partnerships, the state has a window to become a national testbed for resilient, AI-optimized infrastructure—from modular substations to power-as-a-service models.
The question isn’t just how Georgia absorbs global volatility. It’s how fast the state turns that volatility into a platform for strategic leadership.
Arun Rai is Regents’ professor of computer information sciences and Howard S. Starks Distinguished Chair. He is also director of the Center for Digital Innovation.