World Economic Forum maps AI risks for entry-level work

AI entry-level jobs are under fresh scrutiny after the World Economic Forum and PwC published a June report warning that the first step onto the career ladder is being reshaped by artificial intelligence. Its central message: AI exposure can change what workers do, but it does not by itself show that a job will disappear.

The report says 37% of young workers globally are in occupations with medium-to-high exposure to AI-driven task change. That figure climbs to 69% in Northern America, 63% in Europe and 75% in Eastern Asia, according to the Forum.

Those are measures of potential task disruption, not a count of jobs lost. The Forum says entry-level hiring slowdowns are visible, while AI’s specific role in those slowdowns remains disputed.

What could AI mean for entry-level jobs?

The biggest near-term question is whether employers preserve the routes into work for people at the start of their careers as software takes on more tasks traditionally handled by entry-level employees. The Forum says employers should make early-career hiring an explicit part of workforce planning, including targets to maintain or increase intake as they adopt AI.

Its four pressure points are job access, job design, talent pipelines and the fit between education and changing workplace demands. That means companies may need to redesign roles around human-AI collaboration while keeping development and advancement routes in place.

Exposure also varies widely by field. The Forum identifies financial services, information and communication, professional services, science and education as among the sectors with the highest exposure. Agriculture, construction and food services are comparatively less exposed.

Tasks may change before whole occupations do

Research summarized by MIT Sloan points to a less tidy outcome than a simple jobs-versus-machines story. In a study covering AI adoption and task exposure from 2010 through 2023, a role’s share within a company fell by about 14% when AI could carry out most of its component tasks.

When AI affected only some tasks, however, employment in a role could grow as people moved toward work where the technology was less capable, such as critical thinking and generating new ideas. The study found no major AI-driven change in total employment through December 2023, with losses in highly exposed work largely offset elsewhere.

That record has a major time stamp attached: it largely predates widespread use of generative AI. The researchers said it remains unresolved whether earlier patterns will hold as newer tools spread.

Who gets the upside?

Access to AI is uneven, and so may be its benefits. The National Association of Colleges and Employers, citing Gallup, reported that 66% of employees in remote-capable jobs said they used AI tools, compared with 32% of workers in jobs that could not be done remotely. It also said adoption differs across industries, education levels and age groups.

AI could assist human judgment and complex problem-solving, and could potentially help more people take part in specialized work, including coding, legal research and health diagnostics, according to a Northwestern University account of a discussion involving MIT economist David Autor and Microsoft chief scientific officer Eric Horvitz. Those outcomes depend on choices about training, incentives, governance and hiring, the speakers said.

The National Academies’ 2025 study similarly found that AI could raise productivity significantly while leaving many questions about jobs, stability, equity and income inequality unanswered. For the next generation of workers, the first-rung problem is still open: whether employers use AI to narrow the way in, or build new ways to learn and advance.

This story draws on original reporting from CBS News.