AIReshapesCorporateStructuresandHollowsOutManagement,ReportSays

Artificial intelligence is rapidly restructuring the global job market, hollowing out corporate middle management and shifting the employment landscape toward flexible, human-machine collaboration, according to industry leaders and academics gathered at Tsinghua University.

At a June 17 conference launching a new report on skill trends in the AI era — hosted collaboratively by recruitment platform Tongdao Liepin and Tsinghua University's School of Economics and Management — experts warned that AI is no longer a simple technological add-on but an exponential force redefining business models and the nature of work itself.

Tech industry layoffs hollow out the middle

Dai Kebin, chief executive of Tongdao Liepin, highlighted a global trend of “middle-level collapse.” Citing data from tech-recruitment tracker TrueUp, Dai noted that the technology sector has shed roughly 150,000 jobs across 363 layoff events in 2026. May alone saw roughly 40,000 cuts, marking a two-year high.

Companies such as Amazon and Meta have increasingly targeted middle-management roles. Strikingly, AI has become the most frequently cited rationale for these cuts, replacing previous explanations like cost savings or strategic restructuring, even as these tech giants post surging quarterly profits.

The impact is equally severe for entry-level white-collar workers. In China, demand for junior roles in sales, business development, marketing, legal, and procurement has dropped sharply. Even entry-level software developers are seeing severely reduced demand. In Silicon Valley, Dai observed that even top computer science graduates are struggling to find full-time employment, often settling for internships.

“Technology equalization has allowed people who didn't study the major to use the technology to get things done,” Dai said. As tools like Copilot become widespread, the barrier to coding plummets, allowing non-specialists to complete basic development tasks.

This dynamic forces workers to the extremes — either upward into strategic decision-making capacities or outward into skilled blue-collar work. Concurrently, companies are fundamentally shifting their employment models — moving away from long-term, stable hiring approaches toward flexible, task-based project fulfillment.

Organizations shift from headcounts to token costs

The internal structure of the modern enterprise is also facing a radical overhaul. Workers are becoming multi-skilled generalists, and management approaches are shifting from “people managing people” to “people managing agents,” Dai said. Instead of sequential standard operating procedures, workflows are becoming parallel and multi-tasking.

Communication protocols are bypassing middle layers to reach the front lines directly, and corporate cost accounting is transitioning from simple headcount expenses to computing token costs.

To survive, businesses and educational systems must treat AI as a foundational variable rather than an accessory. Yang Bin, director of the leadership research center at Tsinghua’s economics school, said companies must focus on whether their core business structures can withstand and adapt to AI integration.

Tsinghua is actively adapting to this shift by scaling back certain undergraduate majors — despite healthy enrollment figures — and adopting a broader, platform-based educational model. Within a decade, Yang predicted, traditional university curricula will take a back seat to collaborative, extracurricular interactions.

Cultivating meta-capabilities for the AI era

As AI rapidly absorbs technical skills that historically required years of hands-on experience to master, transferrable “meta-capabilities” are becoming paramount for survival.

Guo Xunhua, a professor at Tsinghua, described the new human-AI workplace dynamic as a “guardianship.” Because AI possesses autonomous decision-making and learning capabilities but cannot take legal or ethical responsibility, human workers must specialize in goal setting, strategic guidance, and accountability.

This paradigm shift presents a unique advantage for liberal arts graduates. As natural language effectively becomes the primary programming interface, structural communication skills will become highly valuable, Guo noted.

Ultimately, organizations must move away from viewing AI strictly as a headcount replacement and shift toward embracing human-machine symbiosis. Using technology solely to cut staff is a misguided direction, Guo cautioned, adding that the true historical value of any technology has always been in liberating human creativity rather than replacing humans altogether.