Summary

Headlines about AI and jobs tend to swing between two extremes: mass unemployment or painless productivity. New research from the McKinsey Global Institute (MGI) points to a third outcome that is harder to headline and more important to plan for. AI and related technologies could automate more than half of U.S. work hours over the next decade, yet the United States could still have more jobs in 2035 than it does today.

The problem is not a shortage of work. It is an AI skills gap between the roles that are shrinking and the roles that are growing. MGI estimates that technological and economic change could reduce labor demand by the equivalent of 36 million jobs by 2035, while economic growth could create demand for about 41 million. Roughly 11 million Americans, about 7% of the workforce, may need to move into new occupations to close that distance.

Leaders are not yet organized around this reality. In the EY-Parthenon CEO Outlook Survey of 1,200 chief executives, 72% expect skills shortages to become a bigger barrier to growth than access to capital, yet only 15% are reinvesting productivity gains into workforce development. This article explains what the MGI findings mean in practice, which barriers slow workers down, and how finance and HR leaders can respond before the gap widens.

What Is the AI Skills Gap?

Consider an office and administrative support team at a mid-sized insurer: claims intake clerks, scheduling coordinators, records specialists. MGI estimates that roughly 80% of the hours worked in these occupations could be automated by 2035. That does not mean 80% of those people lose their jobs. It means the work they do today will look very different, and many of them will need new skills to stay employed in the same company or move to a growing field.

The AI skills gap is the mismatch between those two states. On one side sit the capabilities workers currently have. On the other sit the capabilities that growing roles require, from reviewing and validating AI-generated work to redesigning workflows around AI tools. The gap is widest where a worker must change occupations entirely, because the new role often demands skills, credentials, or experience that the old one never developed.

AI Skills Gap vs. Job Displacement

The two ideas are often treated as the same thing, but they point to different responses. Job displacement asks how many positions will disappear. The skills gap asks whether the people affected can reach the positions that are growing. MGI’s research suggests the second question is the more useful one, because net job creation is positive while the path from a declining role to a growing one is where workers get stuck.

This distinction changes how leaders measure success. A company that cuts headcount and reports cost savings has addressed displacement in the narrowest sense. A company that tracks how many employees move into new roles, how many open positions it fills internally, and how workers fare after a transition is addressing the skills gap directly.

What the McKinsey Global Institute Research Shows

The MGI report, titled Workforce in Motion: Skills and Pathways to Future Jobs in the United States, was published in September 2026 and covered by Fortune in October. Its core message is that automation will reshape work faster than it eliminates it. The findings below matter most for leaders who plan workforce capacity, training budgets, and hiring strategy.

Work Hours Absorbed Versus Jobs Lost

The headline figure, more than half of U.S. work hours automatable by 2035, describes tasks rather than people. When automation takes over a portion of a role, the person in that role does not vanish. Their time shifts toward the parts of the job that machines handle poorly, or toward new responsibilities created by the technology itself.

For finance leaders, this reframes the question. The meaningful issue is what the business does with the time AI frees up. If employees use it to improve customer service, develop new ideas, or take on more complex work, the gain compounds. If it simply shortens the time spent on existing tasks, the value tends to evaporate.

36 Million Jobs Reduced, 41 Million Created

MGI’s net estimate is positive: about 36 million jobs’ worth of reduced labor demand against roughly 41 million jobs of new demand, including growth in AI-related industries and services. The United States is also facing an aging population, which means the economy may have more jobs available than workers to fill them. Employers therefore have a strong interest in making it easier for existing workers to move into growing roles.

That interest is practical. Hiring externally for every new skill is slow and expensive, and the labor pool is not expanding quickly. Organizations that build internal pathways gain access to talent that competitors are paying a premium to recruit.

11 Million Workers and 770,000 Moves a Year

About 11 million Americans may need to leave declining occupations for new ones over the next decade. MGI estimates that roughly 770,000 workers a year may need to change occupations, which is about 3.6 times the historical rate. Others will be able to stay in their field as the work evolves around them.

The pace is the challenge. Occupational transitions have always happened, but a rate more than three times higher than normal strains every system that supports them, from community colleges to corporate learning programs to hiring practices.

Office and Administrative Support Roles

Office and administrative support is the most exposed category in the research, with about 80% of current work hours potentially automated by 2035. These roles employ large numbers of people, often with limited formal credentials and modest wages, which makes the transition especially difficult. Workers in these positions frequently hold transferable strengths in coordination, communication, and process knowledge that growing roles value.

The risk for employers is losing that institutional knowledge. A company that treats administrative staff as a cost to be automated may discover that it has discarded the people best positioned to supervise the automated process.

Management, Law, and the Augmentation Effect

In fields such as management and law, where judgment, coordination, and accountability matter, the MGI report finds that AI is more likely to help people do their jobs than replace them outright. It also creates new responsibilities that need training, including checking AI-generated work, ensuring responsible use, and redesigning workflows. These tasks are real, valuable, and not yet part of most job descriptions.

Augmentation still produces a skills gap. A senior associate who once drafted documents now needs to evaluate machine drafts at scale, spot subtle errors, and take accountability for outputs they did not write. That is a different skill set from the one that earned the promotion.

The Credential Barrier

About 85% of growing employment requires a credential or certification. For 38% of those jobs the credential is legally required, and for another 47% employers prefer it. For a worker leaving a declining occupation, a missing certificate can block a path even when the underlying skills are present.

MGI suggests that companies could widen their talent pools by rethinking which qualifications are truly essential and by giving greater weight to proven skills and experience. This is one of the few levers that employers control directly and that costs little to adjust.

How Organizations Can Close the AI Skills Gap

The research points to practical moves that finance, HR, and operations leaders can make now. None of them depends on predicting exactly how AI will develop. All of them improve the odds that people can move to where the work is.

Adopt Skills-Based Hiring

Replacing degree and certificate requirements with skill assessments opens roles to workers whose experience matches the job even if their paperwork does not. Where a credential is legally required, employers can sponsor it. Where it is merely preferred, they can test the underlying ability instead.

Build Internal Mobility Pathways

Internal mobility turns a displacement risk into a talent pipeline. Mapping adjacent roles, identifying the skills that bridge them, and funding the training to cross the gap gives employees a visible future. It also lets the organization fill growing roles faster than an external search would.

Reinvest Productivity Gains in People

The EY-Parthenon survey found that half of CEOs name AI as their biggest source of productivity gains, yet only 15% are directing those gains toward workforce development. A simple policy of earmarking a share of measurable AI savings for reskilling aligns the cost savings with the capability the company will need next.

Measure Transitions, Not Just Savings

MGI recommends tracking whether employees move into new roles, whether companies fill more positions internally, and how workers fare after transitions. These indicators sit next to AI cost savings on the dashboard rather than behind them. Without them, a company can hit its efficiency target while quietly eroding its talent base.

Make AI ROI Visible

Only 16% of CEOs in the EY-Parthenon study report clear, real-time visibility into AI ROI, and 23% say productivity gains are often absorbed before they reach the bottom line. Better measurement helps finance teams see which AI deployments free up time that is redeployed productively and which simply compress existing work. That visibility is a precondition for deciding how much to invest in reskilling.

Challenges and Limitations of Closing the AI Skills Gap

Even well-designed programs run into structural obstacles that leaders should plan for.

• Speed of change: The projected pace of roughly 770,000 occupational moves a year is far above historical norms, which strains training capacity.
• Credential requirements: With 85% of growing employment requiring a credential, many workers face a legal or practical gate that skills alone cannot open.
• Time and cost of retraining: Reskilling takes months or years, and workers in lower-wage roles often cannot afford to pause earnings to do it.
• Unclear ROI: Only 16% of CEOs have real-time visibility into AI returns, which makes it hard to justify workforce investment against other priorities.
• Organizational complexity: Enterprise AI adoption often stalls because of accumulated internal complexity rather than limits of the technology itself.
• Geographic mismatch: Growing jobs are not always located where declining roles are concentrated, so pathways may require relocation or remote work.
• Uneven exposure: Administrative support faces about 80% automation exposure while management and law face augmentation, which means one program will not fit every group.
• Measurement habits: Many organizations still track headcount and cost savings but not transitions, internal fill rates, or worker outcomes.

How to Choose the Right Approach to the AI Skills Gap

The right strategy depends on where an organization is exposed. A company with a large administrative or processing workforce faces high near-term automation of tasks and should prioritize internal mobility and role redesign. A firm dominated by professional roles in management, law, or analysis should prioritize training in AI oversight, validation, and responsible use.

Leaders should also ask what they are actually optimizing. If the goal is short-term cost reduction, hiring freezes and attrition may look efficient. If the goal is a durable talent base, the better measures are internal fill rate, time to proficiency in new roles, and retention of employees who have been reskilled.

Most organizations will combine approaches. Skills-based hiring widens the external pool, internal mobility supports existing staff, and measurement keeps both honest. The common thread is treating workforce transition as a managed program with owners, budgets, and metrics, not as an HR side project.

What Does the AI Skills Gap Mean for CFOs and HR Leaders?

The MGI findings arrived in a newsletter written for finance chiefs for a reason. Capital allocation decisions now determine whether AI productivity turns into growth or leaks away. CFOs who partner with CHROs can set the targets, fund the reskilling, and require the reporting that shows whether people are landing in better roles.

HR leaders carry the operational weight. They need to redesign job architectures, update credential requirements, and build learning programs at a pace the labor market has not seen before. The two functions succeed together or not at all, because workforce capability is both a cost line and a growth constraint.

Choosing the Right AI Skills Strategy for Your Organization

The picture that emerges from the McKinsey Global Institute research is neither alarming nor comfortable. Jobs will exist, and many will be better than the ones they replace, but the route from one to the other is blocked by skills, credentials, time, and cost. The organizations that treat this as a core business problem will have access to talent that others cannot find or afford.

Leading companies are already shifting their questions. They ask not just how much work AI can take on, but what the business does with the time it saves. They track transitions alongside savings, they widen their definition of qualified, and they reinvest a share of productivity gains in the people who will run the next version of the business.

Bronson.AI helps organizations assess where automation will change their workforce, design reskilling and internal mobility programs, and measure the results. If you are planning for the skills your business will need in 2035, Bronson.AI can help you build a pathway that works.