China Sets Out 18 Tasks to Create AI Jobs and Cushion Workers Against Displacement

Employment checks and transition support enter China's AI plans

China has set out 18 tasks across six policy areas to expand employment linked to artificial intelligence while limiting disruption for workers whose jobs could be replaced. The measures include employment risk assessments for some AI projects, advance notice to affected employees and encouragement for companies to explore an unemployment buffer period.

The General Office of the State Council issued the guidelines under document number Guobanfa [2026] No. 29. Dated October 16, 2026, and made public on October 23, the document connects support for AI businesses with worker training, employment services, labor protections and preparations for possible job losses.

Its central approach is to pursue technological progress and employment together. Companies would receive support to develop and adopt AI, while public authorities would consider where new jobs can emerge, which workers need help moving into different roles and how employment risks should affect project reviews.

The policy reaches beyond technology companies. It covers industrial production, agriculture, services, small businesses, schools, recruitment platforms and social insurance. For workers, the most concrete provisions concern notification before large deployments, access to retraining and job recommendations, protection against abusive algorithms and support under existing unemployment insurance rules.

How the policy moves from publication to implementation

The guidelines were issued with State Council approval and address provincial governments, ministries and other central government bodies. Their timetable begins with two dates:

• October 16, 2026: The General Office of the State Council dated the guidelines, numbered Guobanfa [2026] No. 29.
• October 23, 2026: The policy was made public, setting out six areas of action and 18 specific tasks.

Local authorities and relevant departments are instructed to adapt implementation to their circumstances. Human resources and social security departments, together with development and reform departments, are responsible for coordination. The document also requires Communist Party leadership throughout implementation and reporting of major matters to the party leadership and State Council.

The guidelines do not specify a nationwide implementation deadline, a fixed length for the proposed unemployment buffer period, a total funding allocation or a numerical target for jobs created. Their practical effect will depend on how departments and local governments turn the instructions into programs, assessments and services.

AI adoption would face closer scrutiny over jobs

Projects in fields subject to market access reviews would undergo an employment risk assessment alongside their review if their proposed AI applications could cause a substantial employment shock. Market access reviews are checks used to decide whether certain business activities or projects may proceed.

Relevant departments would coordinate support for workers moving between occupations or positions. AI projects in other fields would also be assessed for employment risks according to their circumstances. The guidelines direct innovation resources toward applications with greater potential to create jobs.

Companies are encouraged to explore an unemployment buffer period when deploying AI on a large scale. The document connects that proposal with advance notification for affected employees and assistance with changing jobs or roles. Its wording encourages companies to develop such arrangements, while directing authorities to organize related employment support.

AI deployment would receive priority in jobs where workers are scarce or conditions are dangerous. In industries that employ large numbers of people, the guidelines call for caution over the scope, intensity and pace of adoption to reduce disruption.

Those provisions make the circumstances of deployment central to the policy. Using AI to address a shortage of workers or reduce exposure to hazards receives different treatment from introducing it rapidly into an industry where many people depend on existing jobs. The document also promotes collaboration between people and machines in industrial and agricultural modernization.

Different regions would take different roles

Eastern China is encouraged to develop core AI industries, while central and western regions are guided to take on related activities suited to local conditions, including data annotation and the manufacture of smart devices. Data annotation means labeling material such as text or images so AI systems can learn from it or be evaluated.

The regional approach places employment opportunities across the AI supply chain. Core technology development is one part; preparing data and producing devices are others. The policy seeks new sources of jobs through these connected activities rather than concentrating its employment strategy on AI research alone.

Businesses built around AI and specialized smaller technology companies would be supported. Eligible firms could be included in dedicated lending arrangements intended to help businesses retain workers and expand employment.

Traditional industries also feature in the employment plan. The guidelines call for human and machine collaboration in manufacturing and agriculture, more specialized services for businesses, and more varied and convenient consumer services. Personalized services and spending on experiences are identified as areas where AI applications could support additional employment.

Small businesses would get help adopting AI

Micro, small and medium enterprises would receive support through an AI entrepreneurship program, public services for digital transformation and simpler, less expensive technology products. The aim is to make adoption affordable enough for smaller employers and individual business operators.

The document calls for lawful, equal access to data, algorithms and other inputs needed to use intelligent technologies. It encourages lightweight upgrades, allowing businesses to introduce tools that fit their needs and resources rather than undertaking a costly replacement of their systems.

Two financing tools are named: the National Venture Capital Guidance Fund and the National AI Industry Investment Fund. Government investment funds are expected to support AI business innovation through market based investment.

Open source communities and projects are another part of the plan. Open source software makes code available for others to use and develop under its licensing terms. The guidelines link support for these communities with business creation, alongside startup incubation and assistance for ventures that turn research, vocational skills or demand for services into commercial activity.

For a small employer, these measures address both the cost of obtaining AI tools and the ability to build a business around them. The employment goal is to help firms keep existing workers and gain enough capacity or new business to hire more.

Training would cover students and workers changing careers

Universities, vocational schools and technical colleges are encouraged to adjust their courses and expand subjects urgently needed by AI industries. The guidelines call for more interdisciplinary study and AI micro majors, which are focused programs that can add a specific area of knowledge to a student's main studies.

Technology companies would help develop courses, train teachers and upgrade teaching facilities. Students would gain experience inside businesses, while company engineers would teach in schools. The policy seeks people who can combine AI knowledge with skills from other fields.

A dedicated AI training plan would also serve workers outside formal education. Young people, including university graduates, migrant workers and unemployed people are among the groups encouraged to join employment and entrepreneurship training linked to demand in the AI sector.

Local training should respond to actual market needs. Public training facilities, company training bases and centers for highly skilled workers would provide places to learn. This connects the education measures with the transition support promised to workers affected by automation.

Public access to basic AI education would expand through the National Smart Education Public Service Platform, the Ministry of Industry and Information Technology's online skills learning platform and continuing education networks. The document also calls for new occupational standards and skills certification, naming digital management specialists, AI engineering personnel and AI trainers among the occupations to develop.

AI would also change how people find work and support

Employment agencies would use AI to identify people who may qualify for policies and send them relevant information. Instead of requiring every applicant to discover assistance independently, authorities would analyze the characteristics of service users and direct suitable policies toward them.

The national employment information resource database, the national university student employment service platform and the China Employment public service platform would be upgraded. The guidelines call for better collection and analysis of data to make employment services more precise.

Recruitment models, intelligent interviews and virtual reality training are among the products and services encouraged. AI would also be used in employment policy evaluation, fund management, public consultation and aspects of labor dispute mediation and arbitration.

Workers displaced by AI applications would receive targeted career guidance, training, job recommendations and entrepreneurship support if they meet the relevant conditions. Eligible unemployed people covered by unemployment insurance should receive their benefits promptly and in full under existing rules.

The policy connects those services with the earlier requirements for notification and employment assessments. A worker affected by a major deployment could need several forms of help in sequence: warning that the role will change, an assessment of suitable alternatives, training and assistance finding another position.

Worker protections extend to algorithms, pensions and pay

Employers are instructed to safeguard lawful rights to wages, rest, leave and occupational safety as people work alongside intelligent systems. The guidelines call for stronger consultation and coordination arrangements suited to changing divisions of work between humans and machines.

AI use in human resources management would face action against unlawful practices, including employment discrimination, false recruitment and violations of privacy or rest rights caused by algorithm abuse. An algorithm is a set of computational instructions; in a workplace, it may help screen applicants or organize work.

People in flexible employment and newer forms of work would be encouraged and supported to join the basic pension insurance system for urban employees. The policy also calls for expanding the occupational injury protection pilot for people in new forms of employment.

Pay arrangements form another part of the package. Companies are guided to develop compensation systems suited to AI occupations, improve incentives that reward knowledge and apply reference guidance on minimum wages for different categories of skilled workers.

The document also encourages investment in AI technologies serving the public interest, services and open source communities. Alongside its support for new working arrangements involving intelligent agents, the stated aim is to help more workers increase their capabilities and benefit from technological progress.

Employment monitoring would feed into risk planning

Government departments would share AI industry and employment data and jointly build regional and sector models to detect employment risks. The guidelines seek earlier warnings of disruption, supported by closer analysis of how AI use spreads across industries.

Dedicated surveys would track the creation and replacement of jobs, changes in required skills and the reach of AI applications. Universities and research institutions would provide a base for AI employment research centers, while research projects would examine effects on employment and income distribution.

AI related employment risks would be brought within existing arrangements for preventing and responding to unemployment on a large scale. Authorities would prepare response plans and coordinate assessments and action on labor relations risks.

The package brings investment, deployment reviews, training and social protection into one employment policy. Its central test in practice is concrete: whether businesses can develop useful AI applications while workers receive timely notice, realistic routes into new roles and the protections set out in the guidelines.

Key Points

• The General Office of the State Council set out 18 tasks in six areas, in guidelines dated October 16, 2026, and made public October 23.
• Some AI projects facing market access reviews would undergo employment risk assessments alongside those reviews.
• Companies are encouraged to explore unemployment buffer periods, notify affected workers in advance and support job transitions.
• Regional development, investment funds, smaller business support and new occupational standards are intended to create jobs.
• Training and employment services would help students, migrant workers, unemployed people and workers affected by AI adoption.
• The policy addresses algorithm abuse, pension participation, occupational injury protection and preparations for employment disruption.