AIIsFreezingEntry-LevelHiringMoreThanCuttingJobs,PekingUniversityResearcherSays

Artificial intelligence is reshaping the global labor market not by triggering mass layoffs of existing workers but by causing employers to pull back on hirings for new, entry-level positions, according to a Peking University researcher.

During a recent interview with Caixin on the sidelines of the Summer Davos forum in Dalian, Zhang Dandan, a professor at Peking University’s National School of Development, said that macro employment and unemployment rates remain stable as companies prefer to freeze or reduce external recruitment rather than proactively firing current staff. Even when small-scale layoffs occurred, many affected workers could find new jobs quickly, preventing an extension of an unemployment cycle.

However, the real pressure has fallen heavily on young job seekers as the valuation of specific workplace skills is undergoing a rapid evolution in the era of AI.

Entry-level hiring contraction

While global consensus recognizes that white-collar workers face a higher risk of displacement by AI compared to blue-collar workers, the most visible impact is a steep decline in entry-level job openings.

“The youth labor force in Singapore has already seen a 30% contraction, coming in line with a global trend, with China being no exception,” Zhang said. Many corporate executives interviewed by her team preferred embedding AI into workflows to keep headcount basically unchanged while being cautious about external recruitment, leading to a shrinking of overall hiring quotas.

The rapid advancement of large language models has dealt a blow to certain sectors. For example, coding tools like Claude have impacted entry-level IT positions, such as junior programmers, according to Zhang.

Chinese internet giants have largely avoided mass AI-related layoffs due to concerns over social unrest, but they have dramatically scaled back their annual campus and social recruitment push, maintaining minimal backfilling for natural attrition, Zhang said.

Shifting corporate hierarchy

AI is accelerating a decades-long push toward a flatter organizational structure.

Prior to the widespread adoption of AI, complex business operations required a team of 30 to 40 people. By using AI for data aggregation, basic data analysis and workflow standardization, the same tasks are now performed by a team of 10 to 15 people, Zhang said, citing data gathered from some major internet companies.

Because AI can automate information compilation and cross-tier data transmission, middle-level managers have lost their core functions, weakening hiring demand for such executives, said Zhang, adding that the integration has led to the formation of an ultra-flat management structure where bosses can directly communicate with frontline teams for higher efficiency.

Contact editor Ding Yi (yiding@caixin.com)