The artificial intelligence (AI) race is also very much an infrastructure race. Which country pulls out in front will be largely determined by which country can build powerful data centers at the cheapest, quickest, and safest rates. The United States is increasingly in a technological footrace with China, which is quickly building its own data centers and devoting significant funds to doing so. At the same time, the average U.S. data center takes three years to build.
The formula is simple: To excel in the AI race, American data centers must mitigate costs and maximize output, both in their construction and operation. Or, put differently, efficiency is the name of the game.
That is why trendy Carbon Capture and Storage (CCS) technology is so problematic. Unreliable and inefficient CCS technologies are now being used to power data centers in Europe. It would be a major mistake for our country to import that model here.
Carbon Capture and Storage (CCS) refers to a group of technologies that aim to capture carbon dioxide and store it in a safe place, usually underground, by injecting it into rock. It is supposed to reduce carbon emissions writ large and present an environmentally-friendly option to power large energy projects while remaining commercially viable.
That all sounds good, but the supposed benefits of CCS—despite the alluring claims of its proponents—do not hold up under scrutiny.
For one thing, CCS causes electricity costs in natural gas plants to skyrocket, and general efficiency in natural gas plants to shrink. It is simply not viable in economic terms.
Europe offers a cautionary tale. The European Union (EU) heartily endorsed CCS as a policy tool to harness renewable energy on a broad scale. And the EU and several of its member states put their money where their proverbial mouth is, spending upwards of $20 billion on CCS subsidies. Even so, across the continent, CCS’s economic viability is in serious doubt. And, of course, subsidies are an implicit acknowledgment that CCS projects cannot survive on their own profit margins.
That has not stopped some European energy firms from seeking to adopt CCS-based energy sourcing to power new data centers. And it did not stop the EU from adopting, in 2024, its CCS-focused Industrial Carbon Management Strategy, which will cost European taxpayers an estimated $18 billion through 2050. Once the EU bureaucratic train is out of the station, it is hard to apply the brakes.
The track record of CCS initiatives in the United States further underscores the scheme’s lack of economic viability. When Washington subsidies stop flowing to these facilities, they close down. CCS also, ironically, can carry serious environmental risks.
When carbon dioxide (CO2) is injected into underground rock formations as part of the CCS process, scientists say highly damaging leakage of CO2 back into the atmosphere can occur. Researchers note CCS could cause sea levels to rise dramatically due to thermal expansion from underwater CO2 storage. And Stanford University geologists contend that CCS likely causes seismicity triggers—i.e., earthquakes.
However, that did not stop the Biden administration from pouring money into CCS subsidies. On the basis of an ideological agenda, rather than empirical data, the Biden administration’s Department of Energy allocated a whopping $2.5 billion towards just six CCS projects.
Now, the Trump administration has wisened up to this wastefulness: In 2025, the administration cut nearly $1 billion of funding set aside for CCS initiatives. And yet, some tax credits for CCS remain in place, including those authorized by Section 45Q of the U.S. Tax Code, which distorts economic incentives by giving companies tax credits for CCS projects.
For the United States to triumph in the high-stakes AI and data center races, efficiency and low costs are top priorities. That is why industry must follow the Trump administration’s lead in steering clear of CCS projects which impede America’s ability to affordably and reliably power—and win—the high-stakes global AI race.