Consider this: Generating an image through an artificial intelligence platform requires the same amount of energy as it does to charge your smart phone. For comparison, Google once estimated that it takes just 0.3 watt-hours to process results of a Google search, though that number is likely higher now, since it has integrated generative AI models into even its basic search.

As a result, global energy usage is expected to increase between 13-15% each year over the next decade, requiring an estimated $2 trillion of investment to upgrade utility infrastructure worldwide. With a midsize data center requiring 300,000 gallons of water each day – or enough to supply 1,000 homes – water conservation will also become an increasing priority.

We must all work together to ensure that the AI-powered economy is a sustainable one – and the construction industry will play a critical role in building the infrastructure and facilities that make this future possible.

Here's a look ahead at how AI will impact what we build and how we build it in 2025.  

The growth of AI data centers has significantly impacted the market, even as cloud-based data centers already had a compound annual growth rate of 25%. AI data centers require 10 times the computing density of traditional data centers, with some requiring up to 750MW of energy. That’s enough electricity to power 56,000 homes. 

Finding equipment and materials to build these larger facilities at the speed at which they need to be online remains a challenge. Unlike the supply chain disruptions caused by halted manufacturing during the COVID-19 pandemic, the shortages in data center construction are driven simply by high demand. Essential equipment is being purchased faster than it can be produced, disrupting commissioning and resulting in delayed facility turnover. Prices are also escalating as contractors are willing to pay top dollar for the limited supplies available. Building relationships with suppliers and factoring potential delays into the timeline will be key to project success in 2025 and beyond.

Labor shortages also persist, impacting the construction of data centers. The Associated Builders and Contractors organization forecasts that the construction industry will need to add at least 450,000 workers in 2025 to keep up with demand, a number I expect will continue to grow, as 40% of the current construction workforce is expected to retire by 2035. Compounding the issue is the specialized skillset required for data center construction. Training takes time and, while the industry has ramped up apprenticeship and recruitment efforts, I expect the labor shortage will be a focus for the next several years. This will make retention programs for our existing workforce even more critical.  

Manufacturing growth in 2024 outpaced GDP, fueled by high demand and government investments, with forecasts projecting 4% annual growth driven by policies like the Inflation Reduction Act and “Buy American” policies. The recent CHIPS and Science Act successfully sparked a boom in microchip manufacturing in the U.S., with America’s production of semiconductors expected to triple by 2032. As we build more chip manufacturing facilities, we must also work to mitigate environmental concerns.

A typical microchip manufacturing facility requires up to five million gallons of ultrapure water each day. By the end of the manufacturing process, the wastewater produced is contaminated with toxic solvents and heavy metals, including the per- and polyfluoroalkyl substances known as PFAS. This past summer, the Environmental Protection Agency released its first legally enforceable guidelines on the amount of PFAS allowed in drinking water. While the recently passed Building Chips in America Act exempted some microchip facilities from environmental review, scrutiny of the PFAS used in manufacturing is only expected to grow.

The result is a growing demand for private water construction, where client-owned and operated water treatment systems ensure that manufacturing facilities and data centers comply with local regulations. During the process, water is treated for reuse either as a data center coolant or community irrigation. Private and industrial water projects are expected to make up a significant portion of civil construction projects in the new year.

According to real estate investment firm CBRE, there are 3,871 MW of data center processing capacity currently under construction in the U.S. – up a staggering 69% year over year. The one thing slowing down completion? Limited availability of utility power to energize these facilities. Utility companies are struggling with both generating capacity and transmission to meet these growing power demands. CBRE says the cost of utilizing a U.S. data center is up 20%, due in no small part to the rising cost of powering these facilities.

Constructing renewable energy projects will become critical to mitigating rising energy prices and ensuring there is enough power for the technology of the future. Solar energy has already seen massive expansion in the U.S., representing 65% of new energy capacity added to the grid in the first half of 2024, according to the Solar Energy Industries Association.

Many tech companies also see great potential in nuclear power: Amazon, Google and Microsoft have each announced major investments in nuclear energy to help power artificial intelligence and cloud businesses. Contractors should start preparing now if they plan to expand into nuclear construction: Today’s advanced reactor models are driving the need for new regulatory standards. Industry groups working with the government to update standards to reflect the current technology and accelerate the approval timeline.

The construction industry won’t just build the AI revolution, we’ll be active participants, as we deploy new technology that increases efficiency and enhances safety.

AI is already helping to ensure each worker remains safe on the job. On construction sites, AI systems, including machine learning algorithms and computer vision monitor real-time activities to identify safety hazards, such as workers not wearing proper gear or operating in unsafe conditions.

AI is also transforming the maintenance of construction machinery and equipment through predictive maintenance. Sensors collect data on machinery performance, and AI analyzes this data to predict when maintenance is needed. This approach reduces downtime, prevents costly breakdowns, and extends the life of the equipment. Additionally, AI-powered robotics and autonomous machinery automate tasks such as bricklaying, concrete pouring, and demolition, while drones assist in surveying large areas and monitoring project progress with high precision. These innovations lead to smarter, more efficient construction processes, reduced costs and enhanced safety.

Artificial intelligence has already transformed the way many of us live and work. Over the next several years the construction industry will be kept busy as the world plays catch-up, building the data centers, energy infrastructure and manufacturing facilities that keep the AI economy running.