AI didn’t create the talent shortage. Employers did.

Only 22% of leaders believe entry-level workers are fully prepared for their jobs, and 60% of businesses cite skills gaps as a major barrier to transformation, yet employers have reduced internal training and outsourced workforce development, according to Susan Reichle, president of the International Youth Foundation. Reichle argues that the AI-ready workforce companies imagine hiring largely does not exist, with global demand for AI talent exceeding supply by more than 3 to 1, and that employers must invest in short, practical, scalable training now rather than waiting for others to solve the problem.

I’ve reached an uncomfortable conclusion after a decade helping connect companies with tech talent across Africa, and now leading the International Youth Foundation IYF , a global NGO focused on youth workforce development. I’ve learned that employers have become remarkably good at complaining about talent shortages while investing less and less in talent creation. Only 22% of leaders https://generalassemb.ly/blog/entry-level-workers-still-seen-as-unprepared-soft-skills-gap-widens/ believe today’s entry-level workers are fully prepared for their jobs, and 60% of businesses https://www3.weforum.org/docs/WEF Putting Skills First 2024.pdf say skills gaps in the labor market are among the biggest barriers to transformation. Yet over the past two decades, I’ve seen most employers steadily reducing internal training, raising credential requirements, and outsourcing workforce development to governments, schools, and nonprofits. Then they wonder why the pipeline is dry. This pattern is worth calling out now because AI https://www.fastcompany.com/section/artificial-intelligence is about to make it a lot more expensive. What’s getting lost in the current conversation on AI and jobs is that the AI-ready workforce many companies imagine hiring https://www.fastcompany.com/section/hiring at scale largely doesn’t exist yet. You can’t lay off thousands of workers and replace them with data scientists, AI engineers, and prompt specialists who are still being trained, still scarce, and still concentrated in only a handful of markets. https://www.interface-eu.org/publications/talent-in-talent-out section-1-global-ai-talent-distribution The real constraint facing most employers right now is too few workers with the skills to operate alongside the technology they’re racing to adopt. Recent labor-market analyses estimate that global demand for AI talent exceeds supply by more than 3 to 1 https://divogue.net/ai-engineer-shortage-2026-demand-outpacing-supply/ . And unlike a talent shortage caused by demographics or geography, this is a constraint companies can actually do something about—starting now, before the gap widens further. Some are already figuring this out, and not in the markets you’d expect. In Mexico, IYF partnered with Google.org to build InteligencIA https://iyfglobal.org/press/iyf-equip-60000-mexican-youth-ai-skills-supported-googleorg/ , a program that teaches practical AI skills to young people and government officials in underserved regions. These are places where limited access to technology and digital education already put youth at risk of being left behind as AI reshapes the job market. The curriculum is short by design: 10 hours, mobile-first, self-paced, and built for people who don’t necessarily have a laptop and a quiet private room to learn in. It focuses on responsible, applied AI use rather than theory. More than 15,000 young people have gone through it so far, with a 76% completion rate. Women make up half of participants. These are strong numbers for a program built around a technology that most people assume requires a computer science degree. The plan is to scale to 60,000 by 2027. And take IKEA, which in 2024 set out https://www.raconteur.net/technology/ai-training-employees to train its entire 165,000-person workforce in AI literacy—proof that even companies with no philanthropic angle see this as table stakes, not extra credit. The point is that AI skilling doesn’t have to wait for a perfect curriculum, a four-year degree pipeline, or a generation of new hires. It can start now, with short, practical, scalable training. And it can start with the workforce and communities companies already operate in. That’s a very different posture than the one most employers currently take toward workforce development, which is to treat it as someone else’s job. For decades, the operating assumption has been: Schools prepare workers, governments fund the gaps, and companies hire from whatever pool results. When that pool doesn’t have the skills companies need, the response is usually to complain about the pipeline rather than help build it. That posture was always a little shortsighted. In an AI-driven labor market, it’s no longer viable. If skills gaps are the binding constraint on transformation—and most employers say they are—then closing those gaps isn’t a side project for the philanthropy team. It’s core to whether a company can execute its own AI strategy. The employers that treat workforce development as infrastructure, not charity, will have people ready to work alongside the technology when the rest of the market is still trying to hire its way out of the problem. The companies that succeed in the next decade will be building their own talent—in their own workforce, and in the communities they depend on—while everyone else is still complaining that it doesn’t exist. Christina Sass is president and CEO of the International Youth Foundation.