Skip to main content
A Chinese woman in a tech office, surrounded by computers, looks thoughtfully at a screen displaying AI-related code.

Editorial illustration for AI-Exposed Jobs in China Fell 10 Points Faster, Study Finds

AI-Exposed Jobs in China Fell 10x Faster

• 4 min read

Bloomberg News combed through almost 10 million online job postings in China, stretching from July 2021 to June 2025, and found a pattern that lines up with what economists have warned about for years: jobs that overlap most with AI capabilities are disappearing faster than everything else. The gap is about 10 percentage points, using an exposure index built by researchers at Peking University to rank how much a given role's tasks can be done by AI systems.

The timing matters. Bloomberg set June 2022 as a baseline, just months before OpenAI released ChatGPT and before DeepSeek and ByteDance's Doubao became household names in China's tech sector. Total online job postings fell by half in the three years that followed. Entry-level hiring and creative roles show up as particularly hard hit in the data, according to Bloomberg's reporting.

China's job market was already strained before any of this, dragged down by a slow economy and a property sector that's been in retreat for years. Bloomberg is careful to note that isolating AI's exact contribution is messy, since a general slowdown could hit office-based jobs on its own. Still, the numbers prompted Bloomberg to check whether independent researchers in China were seeing the same thing.

According to a Bloomberg News analysis of nearly 10 million online job postings between July 2021 and June 2025, listings for jobs highly exposed to AI fell about 10 percentage points more than listings for less affected roles.

Why this matters

China's labor data gives us something the US market rarely offers: a large, fast-moving economy where AI's hiring effects show up in hard numbers rather than anecdotes. A 10-point gap between AI-exposed and other job postings, plus Chen Qin's 5.2% net hiring loss estimate, is a concrete signal that the displacement conversation has moved past hypotheticals in at least one major market. For founders building AI tools aimed at entry-level or creative work, this is a preview of what adoption curves can do to demand for junior talent, and it's worth asking whether your product roadmap is accelerating a trend that regulators elsewhere are already nervous about.

For researchers, the involvement of the Chinese Academy of Social Sciences suggests this data will keep getting scrutinized and refined, not dismissed as noise. We'd treat these figures as an early read rather than a settled verdict, China's labor market has its own structural problems that predate AI, but the overlap between a weak economy and AI-exposed job losses is exactly the kind of compounding effect worth tracking closely as similar tools scale in other countries.

Common Questions Answered

What was the key finding from Bloomberg's analysis of Chinese job postings between July 2021 and June 2025?

Bloomberg analyzed nearly 10 million online job postings in China and found that jobs highly exposed to AI capabilities declined approximately 10 percentage points faster than job listings for roles less affected by AI. This significant gap was measured using an exposure index developed by researchers at Peking University to rank how much AI systems could perform tasks associated with different roles.

How did researchers at Peking University measure AI exposure for different job roles in the study?

Peking University researchers created an exposure index that ranks how much a given role's tasks can be performed by AI systems. This index was then applied to the nearly 10 million job postings analyzed by Bloomberg to categorize roles by their level of AI exposure and track their decline rates over the four-year period.

Why is China's labor market data particularly valuable for understanding AI's impact on employment?

China's labor data provides concrete, large-scale evidence of AI's hiring effects through hard numbers rather than anecdotes, which is something the US market rarely offers. As a large, fast-moving economy, China's job posting trends demonstrate that AI-driven job displacement has moved beyond hypothetical discussions and is now showing measurable real-world effects in a major market.

What does Chen Qin's 5.2% net hiring loss estimate indicate about AI's employment impact in China?

Chen Qin's 5.2% net hiring loss estimate, combined with the 10-point gap between AI-exposed and other job postings, serves as a concrete signal that AI-driven job displacement is occurring in China's labor market. Together, these metrics suggest that the conversation about AI's impact on employment has transitioned from theoretical concerns to documented economic effects in at least one major market.

Who would be most affected by the employment trends revealed in this Chinese job posting analysis?

Entry-level and creative workers appear to be most vulnerable based on the study's findings, as these job categories tend to have higher AI exposure. For AI tool founders targeting these sectors, the Chinese labor market trends serve as an early indicator of how their products may reshape hiring patterns and employment opportunities in the coming years.

LIVE19:50METR's AI Graph Update Adds Hard-to-Score New Evaluations