China's State Council has told government agencies to get ahead of AI-driven job losses, issuing national guidelines that turn the employment impact of AI into a standing object of government measurement.

The document, issued on Friday 23 October by the General Office of the State Council and published by Xinhua, sets out 18 tasks across six areas: expanding AI's job-creation potential, making adoption "employment-friendly", raising workers' digital skills, improving employment services, strengthening labour and social security, and monitoring and responding to employment risks.

The operational core is a mechanism to assess employment risks linked to AI adoption, backed by contingency plans. Where an AI project could significantly disrupt the labour market, authorities are told to support job transitions and re-employment rather than leave the adjustment to the market. Displaced workers should receive targeted career guidance, retraining and job referrals.

A second track covers AI pointed at workers rather than used in their place. The guidelines call for regulating AI in human-resources management and for curbing algorithmic discrimination, fraudulent recruitment and privacy violations. They also ask for a rights framework covering human-machine collaboration, social security that reaches flexible and platform workers, and income-distribution rules that adapt as AI spreads.

What the text does not say matters as much. It is not a ban on replacing employees with software — a reading that circulated widely on social media after publication and has no basis in the confirmed text. What is required is that displacement be measured, planned for and cushioned. Nor does it set hiring quotas or name sectors.

Why it matters: in the West, AI governance is argued mostly in the language of safety and fundamental rights — the EU AI Act's risk tiers and documentation duties, prohibitions on specific uses, and in Washington a broadly deregulatory posture. Beijing's frame is labour: an AI policy that reads like an employment policy, with reporting obligations attached.

For firms deploying AI at scale, an employment-risk assessment is a new compliance surface. The guidelines place most of the monitoring and the funding of retraining in the hands of local authorities, so how strictly the assessment is applied will vary by province. The measures also fit a pattern: Beijing has made employment a standing priority as youth unemployment remains politically sensitive, and it is now unusually direct in naming labour-market disruption as a cost of its own industrial push.

Sources