$165.7 billion.
That is more than the gross national product of countries like Kuwait, Slovakia, Bulgaria and Uruguay.
It’s also the projected market size of artificial intelligence (AI) data centers by the year 2034…
That’s an 844.9% increase in market size from this year…
It assumes a compound annual growth rate of 28.3%.
It’s massive growth.
And a bulk of this is happening in the U.S.
That’s because we already have a robust, technologically advanced infrastructure, strong investors in AI-ready data centers and rapid AI adoption in various industries.
U.S.-based companies are increasing their investments in AI-ready data center infrastructure to support emerging technologies, including Big Data analytics, machine learning and generative AI.
Growth in U.S. AI data centers from now to 2034 is projected to be 863.5%, ultimately hitting a $46.1 billion market size. This is faster even than the global growth rate.
However, reaching these projections is proving to be more challenging than we initially thought.
Let me explain…
Data Center Construction Slows
To meet the AI demand, we cannot simply build on existing data centers.
We have to construct new ones … and a lot of them.
Current data centers lack the immense power, cooling and high-performance infrastructure required for AI data loads.
Training large language models for AI requires power-hungry hardware, such as GPUs, which need significantly more electricity and advanced cooling systems than traditional data centers.
Recent data indicate a significant slowdown in the construction of new data centers in the U.S.
In September, there was a 30% year-over-year increase in construction. That’s significantly lower than the 80% increase in December 2024 and January 2025.
According to commercial real estate developer CBRE, North American data center investment volume decreased by more than half in the first half of 2025, reaching less than $1 billion.
One might be led to believe the data center trend was overhyped.
However, that’s not the case.
The primary reasons for the slowdown in data center construction can be traced back to three main issues:
- Economic uncertainty — Inflation is still higher than the 2% range set by the Federal Reserve, and growth is not accelerating as rapidly as once thought.
- Geopolitical conflicts — Material sourced for these new data centers comes from everywhere. The U.S. is ramping up its trade war with China, which is a source of some of that material. Until the U.S. can meet 100% of its data center needs domestically, global issues will continue to have an impact.
- Power supply challenges — As mentioned above, these new data centers require an immense amount of electricity to function. This puts a massive strain on the existing power infrastructure. Until that infrastructure is either upgraded or bypassed with a new power supply, this will be the biggest challenge for new data center construction.
The first and second issues are not easily solved. However, the power supply problem is being actively addressed by both the public and private sectors.
The U.S. government has selected four sites: the Idaho National Laboratory, the Oak Ridge Reservation, the Paducah Gaseous Diffusion Plant, and the Savannah River Site to develop energy generation projects.
Companies with a vested interest in advancing AI are also investing millions in solving the power problem.
So, while data center construction is slowing, interest isn’t waning. There are significant opportunities to invest in companies working to grow the industry and reach the $46 billion projection.
That’s all from me today.
Safe trading,
Matt Clark, CMSA®
Chief Research Analyst, Money & Markets