With $580B projected for AI data center spending in 2025—more than new oil development—governments and Big Tech must now grapple with infrastructure demands, grid strain, and public accountability.
A Spending Shift Signals a New Resource Race
The International Energy Agency (IEA) reports that $580 billion will be invested globally in AI data centers in 2025—$40 billion more than what will be spent on new oil supplies.
- This symbolic tipping point suggests a fundamental economic shift: data is the new oil, and compute power is the new engine of growth.
- But unlike oil, data centers aren’t just energy consumers—they are energy vortexes, demanding vast resources, water, and land.
AI’s Unquenchable Thirst for Energy
AI workloads are incredibly power-hungry. A single data center can consume as much electricity as a small city.
- Generative AI models require massive GPUs running 24/7.
- As these systems scale, the energy grid is increasingly strained—especially in regions already operating at capacity.
- Countries like the U.S., Ireland, and Singapore are now debating moratoriums or stricter approvals for new centers.
IEA forecasts that by 2026, data centers could consume more than 1,000 TWh of electricity per year—comparable to the annual usage of Japan.
Climate Tech’s Paradox: Emissions vs. Optimization
AI is also touted as a climate solution, helping optimize energy systems, model carbon capture, or improve agriculture. Yet:
- Most data centers still rely on fossil-powered grids, and cooling systems often waste millions of gallons of water annually.
- If not properly managed, data centers could cancel out the emissions savings AI is meant to enable.
- The sector is now under pressure to adopt green energy commitments—and fast.
Should Taxpayers Foot the Bill for Big Tech’s Infrastructure?
With costs soaring, Big Tech is increasingly leaning on governments to help fund data center development.
- OpenAI’s CFO recently walked back claims that the company was lobbying for “government backstops” for loans, but confirmed it’s seeking support through programs like the CHIPS Act.
- This raises uncomfortable questions:
- Should taxpayers subsidize infrastructure for trillion-dollar companies?
- Or should those companies be held accountable for environmental and social impacts?
Public funding without public benefit risks deepening digital inequality and creating privately owned critical infrastructure that shapes everything from health to education.
A Broader AI Infrastructure Boom
Other major developments shaping the AI infrastructure race include:
- Israeli startup Wonderful raised $100M Series A to scale AI agents for customer service, a sector now viewed as AI’s “killer app.”
- Einride, the Swedish electric truck and autonomous pod company, just announced a SPAC deal, highlighting how AI and clean transport are converging—albeit with skepticism about profitability.
- Governments are racing to regulate AI, even as they invest billions into its foundations—creating a feedback loop of innovation, exploitation, and oversight.
Risks Beyond Energy: Surveillance, Sabotage, and Spyware
The infrastructure that powers AI isn’t just about clean data—it’s also a target.
- China-backed groups like Salt Typhoon are reportedly “pre-positioning for sabotage” in global infrastructure systems, raising cybersecurity red flags.
- Meanwhile, government-grade spyware is being used to track journalists and activists, with mobile design flaws making detection nearly impossible.
- As AI and state power intertwine, infrastructure security becomes not just a tech issue, but a geopolitical one.
Are Data Centers the New Oil Fields? Not Quite—But They’re Close
The comparison is apt in terms of value, scale, and centrality to economic power, but it falls short in one critical way:
- Unlike oil, data centers can serve as platforms for solutions—but only if built sustainably, transparently, and with broad public benefit in mind.
- Without intervention, we risk repeating the environmental and geopolitical mistakes of the fossil fuel era.
Data centers are now drawing more investment than oil, with $580B expected in 2025—driven largely by AI demand. While they offer transformative potential, they also pose serious environmental, energy, and ethical challenges. As governments consider subsidizing infrastructure, the true cost of AI’s growth is coming into sharp focus.
