Artificial intelligence could force about 11 million American workers, or roughly 7% of the entire U.S. workforce, to change professions over the next decade. An even larger process will unfold in parallel: about 70% of employees will have to adapt their work in one way or another to a new reality, even if their profession does not formally disappear.

This is the conclusion reached by the McKinsey Global Institute. Researchers believe that the transformation of the labor market under the influence of AI could happen three to four times faster than major changes caused by previous technological shifts.

This does not mean that all 11 million people will necessarily lose their jobs. The main problem may be more complex: jobs will remain, but people will have to move between professions much faster and learn new ways of working, Axios.com reports.

Transitions between professions will accelerate severalfold

According to McKinsey, on average about 770,000 Americans per year may be forced to switch to another profession over the next decade.

For comparison, the long-term average is about 215,000 people per year.

A similar scale was already seen during the COVID-19 pandemic, when due to sharp changes in the economy and demand structure, this figure reached about 788,000 people per year.

But now it is not about a temporary shock. Researchers expect that the AI-driven restructuring will continue for many years and affect a much wider range of professions.

At the same time, changing professions is only one part of what is happening.

Most workers will not have to leave their profession

For about 70% of American workers, changes will occur directly within their existing profession.

AI may take over some tasks, change the sequence of work, or require people to master new skills. As a result, the position itself will remain, but its content will become different.

McKinsey compares this process to previous general-purpose technologies. The internet did not destroy most professions, but it radically changed how people did their work. The same happened with other major technological shifts.

The difference, according to the researchers, lies primarily in the speed and scale.

“We will see new jobs and new roles emerge, and existing roles will be reshaped—just as happened with other general-purpose technologies in the past,” says Anna Curtis, a McKinsey partner and one of the study’s authors. But, she says, this time the changes will happen faster and affect more people.

The biggest problem may arise after specific tasks disappear

At first glance, McKinsey’s figures do not look like a forecast of mass unemployment. If a significant share of professions remains and some tasks are replaced by others, the labor market as a whole may continue creating jobs.

But for a specific person, the transition may prove much more difficult.

A worker whose profession has changed because of AI will not necessarily be able to immediately move into a new position. A gap may arise between disappearing or shrinking tasks and growing demand for other skills.

That is why the authors of the analysis see the main risk not necessarily in the complete disappearance of jobs, but in the difficulty of transitioning between them.

The economy may have enough jobs overall, but millions of people will find it hard to get the specific jobs that will be in demand.

Anxiety is already showing up in workers’ sentiment

Expectations of these changes are already affecting how Americans feel about work.

According to Glassdoor, in September the employee confidence index reached a new low. Mentions of artificial intelligence in employee reviews rose by 164% compared with last year.

This does not mean that all these mentions are directly related to job losses. But AI is becoming an increasingly visible part of workers’ conversations about the future of their profession and company.

Historically, major technological changes have extended far beyond businesses themselves. For example, the Industrial Revolution was accompanied by labor conflicts and the rise of the trade union movement.

Therefore, the consequences of the current AI transformation may show up not only in employment statistics, but also in politics, culture, and relations between workers and employers.

Mass disappearance of office professions is still a long way off

Back in 2025, Anthropic CEO Dario Amodei predicted that AI could eliminate half of all entry-level jobs for knowledge workers within one to five years, and unemployment could rise to 10–20%.

So far, that scenario has not materialized on a broad scale.

The U.S. unemployment rate stood at 4.1% in August, meaning there are still no signs of the predicted sharp collapse in employment.

But that does not mean AI has had no impact. A more gradual restructuring of the labor market may be less visible in overall statistics while being more painful for certain groups of workers.

The key question is becoming not only how many jobs there will be, but also what skills will be required to get them.

The main question is whether people will manage to adapt

The picture emerging now differs noticeably from the simple scenario that “AI will take jobs away from people.”

A much more complex process appears more likely: some professions will shrink, new roles will emerge, and a huge number of existing positions will change from within.

For the economy, this may mean not the disappearance of work as such, but a large-scale restructuring of who does what, where, and exactly what kind of work.

And here, speed becomes the key factor. If technological changes really do happen three to four times faster than the historical norm, workers will have significantly less time to transition from old skills to new ones.

As a result, the main challenge of the AI era may be not a situation in which people simply have nothing to do, but one in which work exists—but millions of people will have to relearn how to get it and how to do it.