How the AI talent race is reshaping recruitment.
A new survey of 250 technical leaders reveals a striking paradox: Companies are dramatically increasing AI investments—some by up to 75 percent in 2025—while simultaneously finding a talent well that is running dry. Ninety-four percent of tech leaders identify talent shortages as their primary barrier to AI innovation, and 85 percent are pausing or stretching out timelines because they can’t find the people to build and tailor the technology.
As in other cases of extreme shortage, technical leaders are relying on the “spot” market to fill gaps. Rather than hiring full-time AI technical experts—which can require up to four months in the highly competitive market for top engineering talent—companies are creating “blended teams” that integrate contract or gig workers with permanent staff. Businesses are buying fractional time rather than adding to headcount. The surveyed tech leads report this strategy significantly improved their companies ability to innovate, create successful final products, and speed up delivery to customers. Freelancers now make up more than 25 percent of product teams in over half of the surveyed organizations, and 92 percent of the firms expect to increase freelance engagements in the next 24 months.
The investment priorities for 2025 reflect the bias toward “just-in-time” talent. The firms surveyed are beginning the transition from exploration of large language models to the more difficult job of integrating LLMs into business processes and systems: developing platforms, monitoring tools, and infrastructure models that can be scaled to meet new needs. Encouragingly, the survey also finds that technical leaders are looking to AI systems to generate new revenue rather than simply reducing costs by substituting technology for human workers.
Of course, there’s another potential solution to the talent shortage: reskilling. Coders are increasingly finding basic tasks automated by—wait for it—generative AI. These workers’ intrinsic bent toward technology and background in computer programming ought to provide a foundation for many struggling coders to retool as AI “whisperers” for businesses of all sorts, not just the tech sector itself. Over time, these reskilled coders could be upskilled to fill more of the technical expert roles the market currently needs. Entrepreneurial workforce development agencies, community colleges, and other higher ed institutions might consider this a new angle for remaining relevant in the age of AI.
