Trump AI Data Center Delays: Key Obstacles Unaddressed
▼ Summary
– Donald Trump prioritized rapid AI data center construction to compete with China, but his tariffs on Chinese imports are hindering these projects.
– Nearly half of planned US data centers this year face delays or cancellations due to shortages of key power infrastructure components like transformers and switchgear.
– Wait times for these components, historically sourced from China, have increased from about two years to up to five years, potentially impacting the US lead in the AI race.
– US manufacturing cannot meet the demand for this equipment, forcing many firms to accept tariffs and security risks to import from China for faster delivery.
– Trump has not addressed this infrastructure bottleneck, even as data center builders note that securing power is futile without the equipment to use it.
A major push to rapidly construct AI data centers across the United States is encountering severe logistical delays, directly challenging a key presidential priority. Last year, a series of executive orders identified the swift build-out of these facilities as a top national objective to maintain a competitive edge in the global AI race. Current reports, however, indicate that the administration’s own policy tools are creating a critical bottleneck.
The aggressive imposition of tariffs on Chinese imports is now cited as a primary cause for widespread project slowdowns. Industry analysis reveals that nearly half of the data centers planned for activation this year face postponement or cancellation. The core issue is a shortage of essential electrical components, including transformers, switchgear, and batteries, which are fundamental for building the power infrastructure every facility requires. For decades, the U. S. has largely depended on Chinese manufacturing for these parts. Where lead times were once two to two and a half years, suppliers now quote waits of up to five years.
This timing is particularly consequential given estimates that China trails the U. S. in AI development by approximately the same five-year period. The stated goal is to reduce this foreign dependency and bolster domestic manufacturing capacity. Presently, however, American factories cannot meet the soaring demand from the AI sector, leaving developers in a bind.
Market intelligence indicates that only about one-third of the largest AI data centers slated to begin operations in 2026 are actually under construction today. Faced with these constraints, many companies report feeling hamstrung. Their urgency is leading some to accept the financial burden of tariffs and navigate potential national security risks in an attempt to secure Chinese equipment on accelerated schedules.
This material challenge appears at odds with recent directives from the White House. In March, an executive order commanded technology firms to “build, bring, or buy” the necessary power for their data centers, yet it did not address the underlying scarcity of the hardware required to utilize that power. Builders are left confronting a stark reality: the source of electricity is irrelevant if the equipment to connect to the grid remains unavailable. The success of the national AI infrastructure ambition now hinges on resolving this fundamental supply chain disconnect.
(Source: Ars Technica)
