China Pushes AI Literacy Into Schools as Nearly Half of Provinces Set Binding Plans

A Reuters review found Shanghai mandating 30 hours of AI teaching a year and Ningxia and Gansu adding AI literacy to entrance exams by 2027.

WHY IT MATTERS

Local governments now buy AI classroom equipment, training and platforms without a national standard for what pupils should learn, so suppliers must meet differing provincial rules.

Nearly half of China’s provinces and municipalities now have binding plans to teach artificial intelligence in schools, according to a Reuters investigation published on November 16. The goal is universal AI literacy among the country’s almost 200 million primary and secondary pupils by 2030. Reuters reviewed provincial blueprints and more than 100 academic papers, visited schools and interviewed more than a dozen educators, parents and experts.

The local rules are specific. Shanghai requires at least 30 hours of AI teaching a year in every grade. Beijing plans 100 AI pilot demonstration schools, and Ningxia and Gansu intend to put AI literacy into high-stakes school entrance exams by 2027.

What China lacks is a shared definition of the skill it is trying to teach. Reuters found little agreement on what AI literacy means or how to measure it, and the Ministry of Education did not respond to its request for comment. "Implementation is all over the place," Ted Mo Chen, chief of staff at the education technology company ClassIn, told the agency.

National targets set the date and leave the detail to provinces

The 2030 deadline comes from Beijing. The AI+Education Action Plan, issued by the Ministry of Education with the National Development and Reform Commission, the Ministry of Industry and Information Technology, the Ministry of Science and Technology and the National Data Administration under a document dated April 2, 2026, calls for an AI education system spanning every school stage and the wider public by 2030. It tells the ministry to keep improving its general AI education guidelines for schools, to bring AI into local curricula and to have regions set course objectives, content and teaching hours for each stage. It also asks for teacher AI literacy standards and for AI to be added to teacher qualification exams.

An earlier ministry notice, reproduced by the Shanghai Municipal Education Commission in December 2024, set out the progression: younger primary pupils experience AI, older primary and junior high pupils learn to understand and apply it, and senior high students build projects. It called for more support for rural and remote schools.

Some provinces have gone beyond teaching. Reuters reported that Guizhou is adding facial recognition and behavior analysis to classrooms, Shanghai is deploying health monitoring systems, and Yunnan plans to have AI record and analyze lessons in 14,000 classrooms.

Spending is rising faster than the evidence on learning

Reuters counted more than 100 government procurement documents committing at least US$31 million in the past year to AI classroom equipment, training and platforms. Reuters said the true total is likely higher. The consultancy iResearch, cited by Reuters, estimates that 2025 spending on education digitalization came to just under US$83 billion, and that generative AI made up 46% of AI education procurement.

The research on outcomes is mixed. One study of nearly 27,000 Chinese teenagers linked AI adoption to an 18% rise in homework scores but a 20% fall in monthly exam scores within six months, with the steepest drops in social sciences. A survey of more than 320,000 pupils found that those using AI at least three times a week scored lower on memory, comprehension, analysis and creativity. Neither study establishes that AI caused the differences.

"Mature, widely adopted assessment rubrics for compulsory school-level AI competency are still being built," said Selena Guo, portfolio lead at the research firm China Policy. The Next Web, which summarized the investigation, set it against Stanford’s 2026 AI Index finding that the United States relies on nonbinding guidance without dedicated funding.

Rural schools are testing AI as a substitute for missing teachers

Reuters contrasted two schools. Beijing New Talent Academy, a private boarding school charging US$39,000 a year, teaches AI, robotics and programming; one 13-year-old pupil, Peng Gaodi, has built more than 200 apps with AI. In rural Anhui, Funan County Qinqing High School, whose 4,000 pupils are mostly children of migrant workers, uses AI-generated lessons in core subjects to cover a shortage of qualified teachers, drawing on models from Alibaba’s Qwen, ByteDance and Z.AI.

Headmaster Yang Tianliang credits the system with lifting university admission rates from below 60% to 90%, a claim Reuters could not verify, and estimates that at least 200 other schools run similar setups. EastFrontier has covered earlier classroom products, such as iFLYTEK’s AI blackboard.

The 2027 entrance exams in Ningxia and Gansu are the first point where AI literacy will carry real stakes for pupils. Whether the ministry publishes a national assessment framework before then will decide whether provinces end up testing the same skills or each set their own.