The skills gap is one of the most frequently cited challenges in talent acquisition — and one of the most frequently misunderstood. The aggregate number used most often, from research compiled by Truffle's talent acquisition statistics, is that 75 percent of employers worldwide cannot find the talent they need for open roles. That number is real, but treating it as a single problem leads to interventions that are misdirected. The skills gap is actually a set of overlapping structural deficits, each driven by different causes and requiring different solutions.
Gap One: The AI Skills Deficit
The most acute and fastest-growing component of the skills gap is the demand-supply imbalance in AI-related capabilities. Stanford HAI's 2025 AI Index documents that AI skill demand in U.S. job postings grew 20 percent between 2023 and 2024. More dramatically, demand for generative AI skills specifically grew by a factor of four year-over-year, with over 66,000 job postings mentioning generative AI in 2024 versus 16,000 in 2023.
The supply side of this market is growing — new AI PhDs increased 22 percent from 2022 to 2024 — but the pipeline of formal credentials cannot keep pace with demand that is growing at 4x annually. SHRM research shows that 83 percent of HR leaders recognize the need for new AI-related skills in their workforce. Only 55 percent of workers at organizations that invest in upskilling use AI daily — compared to 27 percent at organizations that do not. The upskilling investment gap is as important as the hiring challenge.
Gap Two: The Experience-Credential Mismatch
A second dimension of the skills gap is structural rather than supply-side: it is the mismatch between what organizations require on paper (credentials, degrees, specific years of experience) and what actually predicts success in the role. This is the gap that skills-based hiring addresses, and the data on its scale is sobering.
Research on skills-based hiring outcomes — referenced by IBM, LinkedIn, and multiple academic analyses — consistently shows that degree requirements filter out 20–40 percent of candidates who can demonstrably perform the role, while admitting a significant percentage of degree-holding candidates who cannot. The credential requirement functions as a noisy proxy for the actual skill, producing both Type I errors (qualified candidates excluded) and Type II errors (unqualified candidates included) that do not appear in most organizations' hiring analytics because quality of hire is rarely attributed back to screening criteria.
The organizations making the most progress on this dimension of the gap are those that have replaced credential requirements with skills assessments for roles where the relevant capabilities are testable: software development, data analysis, sales, financial analysis, and operations management. Lightcast's analysis of U.S. job market data for Stanford HAI shows that AI skill clusters have surpassed machine learning as the most-sought-after skill cluster in U.S. job postings — but that organizations asking for "artificial intelligence" as a skill are competing for a fraction of the people who have it, because credential-based filtering excludes the self-taught and experience-developed majority.
Gap Three: The Geographic-Demographic Mismatch
A third component of the skills gap is geographic and demographic: the workers with the skills organizations need are not always in the places organizations are hiring, and the pipelines from which organizations traditionally recruit do not reflect the full distribution of available talent. Remote work has partially addressed the geographic mismatch — organizations willing to hire across geographies access a substantially larger effective talent pool. But the remote-eligible share of roles is still concentrated in knowledge-worker categories, leaving the geographic constraint intact for manufacturing, healthcare, and operational roles.
The demographic dimension is more complex. SHRM's State of AI in HR 2026 research shows that AI job displacement risk is not distributed evenly across demographic groups — certain roles with high automation exposure are disproportionately held by specific demographic segments, while the emerging AI-adjacent roles that are growing fastest require credentials and backgrounds that reflect historical inequities in educational access. Talent leaders who address this dimension through intentional pipeline development, apprenticeship programs, and targeted upskilling investments are both addressing the skills gap and building the diversity of their talent function.
"The 75% of employers who say they can't find qualified candidates aren't all facing the same problem. Some face genuine supply constraints in scarce skill categories. Some face self-imposed constraints from credential requirements that are narrowing their own pool. Diagnosing which gap you're in is the prerequisite for addressing it."
Gap Four: The Reskilling Velocity Problem
The World Economic Forum's Future of Jobs Report 2023 identified that 44 percent of workers' core skills would be disrupted in the next five years. The 2025 WEF follow-up has not reduced that estimate. The reskilling challenge — moving a large share of the existing workforce from skills that are declining in value to skills that are increasing in demand — is a problem of velocity. Training programs take months to years. Job transitions take months to years. The AI capability curve is moving faster than the human development infrastructure can track.
Organizations that are building real competitive advantage on the reskilling dimension are those treating internal skill development as a talent acquisition strategy, not just an HR function. When internal mobility and upskilling can fill roles that would otherwise require external hiring in constrained markets, the organization builds resilience against the structural skills gap rather than simply bidding against competitors for scarce external talent.
What the Data Suggests for Talent Strategy
The multidimensional nature of the skills gap requires a multidimensional response. For AI-skill-specific gaps: invest in targeted upskilling programs, partner with universities and bootcamps at the apprenticeship level, and design roles that pair high-AI-skill profiles with domain expertise rather than requiring both from one person. For credential-mismatch gaps: implement skills-based hiring for roles where capabilities are testable, and measure quality-of-hire by screening criteria to build the evidence base for ongoing policy evolution. For geographic-demographic gaps: expand remote eligibility for appropriate roles and build proactive partnerships with organizations working on historically underrepresented talent pipelines.
The takeaway: The skills gap is real, growing, and consequential — but it is not a single thing, and it does not have a single solution. Diagnosing which of the four dimensions is most acute for your organization's specific hiring portfolio is the prerequisite for building a strategy that actually moves the needle. The organizations that do this analysis are making progress. The ones treating the 75 percent statistic as a single problem are spinning in place.