AI Data Center Boom Faces Reality Check As Many Projects Risk Never Being Built
Many announced AI data center projects may never be built due to power shortages, local opposition, and permitting delays, according to experts on TechCrunch's Equity podcast.
The ongoing expansion of artificial intelligence infrastructure has fueled a massive wave of data center announcements, yet a substantial portion of these proposed projects may ultimately never be constructed. While physical and regulatory hurdles—such as intense local community opposition, long permitting delays, and severe competition for electricity—are already creating bottlenecks, an even more fundamental issue is emerging. A significant share of the projects being publicized across the industry are purely speculative, raising questions about how much capacity will actually come online to support the billions of dollars currently flowing into AI technology.
Dissecting speculative compute infrastructure
To evaluate which data center projects are likely to materialize and which will stall before breaking ground, TechCrunch explored the topic on its flagship podcast, Equity. Hosts Rebecca Bellan and Tim De Chant spoke with Brent Nelson, the senior managing director of markets and strategy at Ascend Analytics, to break down the realities of data center planning.
The discussion focused on separating ground-ready infrastructure proposals from high-risk projects that face systemic barriers. Bellan, a senior reporter at TechCrunch who covers the business, policy, and emerging trends shaping artificial intelligence, alongside De Chant, a senior climate reporter at TechCrunch and lecturer in MIT’s Graduate Program in Science Writing, examined how power availability and regional planning dictate the future of compute capacity.
With billions of dollars in AI infrastructure investments on the line, understanding the gap between announced projects and actual build-outs is crucial for the industry. The podcast episode, produced by TechCrunch audio producer Theresa Loconsolo, highlights how energy access and local governance are fast becoming the ultimate gatekeepers of artificial intelligence scaling.
Obstacles in infrastructure development
The friction surrounding data center development extends beyond balance sheets and corporate announcements. Local opposition across various regions has intensified as communities raise concerns over land use and resource consumption. Simultaneously, permitting processes create extended delays, requiring extensive environmental and municipal reviews before construction can begin.
Compounding these regulatory and social challenges is a critical scramble for electricity. Modern AI workloads require unprecedented amounts of power, forcing developers to compete for access to grid connections that are often constrained or slow to expand. De Chant, who holds a PhD in environmental science, policy, and management from the University of California, Berkeley, and was a Knight Science Journalism Fellow at MIT, has tracked these intersecting climate and technology challenges closely.
Because many announced facilities fail to secure necessary power agreements or local approvals early in the process, industry estimates regarding available compute power can easily become inflated. For organizations monitoring the expansion of AI capabilities, distinguishing between actionable supply and speculative planning is essential.
What it means for developers
For software engineering teams and technological leaders, the gap between announced and actual data center capacity highlights the practical challenges of scaling AI-powered applications. When infrastructure growth stalls due to grid limitations, permitting bottlenecks, and local pushback, backend compute costs and raw model availability can become unpredictable.
Developers relying on large AI models need reliable ways to build and deploy applications without worrying about single-provider capacity issues or fluctuating infrastructure constraints. To mitigate these uncertainties, developers can try top AI models cheaply through one API at https://apixoai.online.
As billions of dollars remain tied up in speculative developments, software teams must stay grounded in infrastructure realities. Tracking whether physical data centers actually obtain electrical capacity and municipal permits provides a far clearer picture of future compute availability than relying solely on press releases and investment figures.
Source: Why the data center you dread might never even be built — TechCrunch AI. Written by the Apixo team from that report.
One key for Claude, GPT, GLM, DeepSeek and more. Pay per token with crypto.
Get your API key
