A randomized field experiment conducted in Guangzhou (n=758, December 2024 to January 2025) found that white-collar workers given information about LLM occupational-exposure rates became significantly more likely to perceive their jobs as at risk, reduced support for government AI-industry subsidies, and increased support for unemployment insurance, universal basic income, and retraining programs. Notably, the treatment did not shift preferences toward upskilling or STEM education, but it did raise the share of respondents wanting their children to pursue public-sector careers, primarily for job stability. For financial institutions, this evidence from China's largest economy suggests that perceived AI displacement risk may generate political headwinds for AI industrial policy and measurable shifts in household career and savings preferences, with implications for modeling labor-market sentiment and regulatory trajectory.
438 other tracked developments touching the same claims.