The disruption appears in tasks, pay, and hiring

The reported experiences do not form a national causal estimate, but they reveal several channels at once: teams shrink, routine creative and technical tasks become easier to automate, training work appears temporarily, and independent workers use the same tools that weaken established roles.

The rapid increase in enterprise adoption suggests these are not isolated experiments. As models and agents move into production, worker bargaining power can change before official employment categories show a clean displacement signal.

Do not confuse adoption with shared prosperity

China also faces a housing downturn, weak consumption, record numbers of graduates, and long-term population aging. Those forces can amplify or mask AI’s labor effects, so policy should avoid both technological fatalism and premature reassurance.

Governments and employers should publish wage changes, hiring at entry and mid-career levels, transition duration, training completion, placement quality, and income support. Productivity becomes socially durable only when people can see a credible route through the change.

Primary trail

Go to the source

Read the evidence behind this analysis. External links open in a new tab.

Associated Press — Chinese workers adapt as AI transforms jobs