America's AI infrastructure boom is generating an unprecedented labor crisis, and the companies feeling it most aren't Amazon or Microsoft. They're the mid-sized firms trying to compete for the same talent pool with a fraction of the resources.
Following the U.S. announcement of a $33.3 billion public-private partnership between SoftBank and the Department of Energy to build a major AI data center site, demand for engineers, data scientists, and infrastructure specialists has accelerated well beyond Silicon Valley. A March 2026 Deloitte report found that data centers and power companies are now competing for the same pool of AI and computer specialists, even as power demand is expected to rise to 176 gigawatts by 2035.
For growth-stage companies, the challenge goes beyond just salary. Large enterprises offer prestige, massive datasets, compute infrastructure, and the chance to work on globally visible products. These advantages carry as much weight as compensation for many engineers. Add dedicated university pipelines and global recruitment teams, and mid-sized firms are often left hiring reactively and under pressure.
But Nisum's Global AI Leader Guillermo Delgado argues that mid-sized companies may be approaching the problem from the wrong angle.
"The key is how much intelligence each engineer can leverage," Delgado said. Rather than trying to outbid Big Tech for headcount, he argues that companies should focus on how effectively their teams can use AI across workflows like testing, quality assurance, and code migration.
Done right, AI-integrated development environments become a structural advantage. Now, the true competitive divide is defined less by a firm's total engineering headcount and more by the enhanced capabilities and output of every individual hire.
The stakes extend beyond individual companies. With 70% of U.S. graduate students in AI-related fields being foreign nationals, and recent DHS proposals creating new uncertainty for international STEM graduates, the window for accessing global talent is narrowing. In an interview for Unite AI, Delgado also flagged that as AI ecosystems grow in regions like Latin America, the gap between building in the U.S. and abroad is beginning to close.
"The U.S. risks losing not only engineers but also the knowledge transfer and institutional investments that follow them," he said.
This article originally appeared on Unite.AI.
