China Wants Every Schoolchild to Be AI-Literate by 2030, but Early Data Raises Hard Questions

China is running one of the largest classroom experiments in history. Beijing wants all of its schoolchildren to be AI-literate by 2030 and is spending heavily on tools, training and mandatory lessons. Early research, however, shows mixed results. One large study found students’ exam scores dropping after they started using AI. Rights concerns also grow as some regions plan to use AI to watch students in the classroom.

Nov 16, 2026 - 10:54
0
China Wants Every Schoolchild to Be AI-Literate by 2030, but Early Data Raises Hard Questions

.

A State-Driven Race for AI Talent

With almost 200 million students in primary and secondary school, China is pushing artificial intelligence into its classrooms on a scale no other country has tried. The government sees AI as the key to its economic future and its technology rivalry with the United States. It hopes early exposure will produce the homegrown innovators needed for major breakthroughs.

The effort is centrally directed. In May 2025, China's Ministry of Education issued two guidelines. One sets out a step-by-step AI curriculum for all school levels, and the other regulates how generative AI (tools that create text, images or code on request) may be used. Official rules bar primary pupils from using such tools on their own. They also forbid students from handing in AI-generated content as homework or exam answers, and they bar teachers from letting AI replace their core teaching role.

Local governments have gone further. According to Reuters, nearly half of China's provinces and cities now have binding AI-in-education plans. Shanghai requires at least 30 hours of AI instruction per grade each year. Ningxia and Gansu plan to make AI literacy part of high-stakes school entrance exams by 2027. Beijing city officials say more than 1,400 schools in the capital already teach AI courses to roughly 1.83 million students.

By comparison, the United States relies on decentralized, non-binding guidance. Stanford University's 2026 AI Index found it has no dedicated funding for the purpose and no state-level teacher standards.

.

Two Very Different Classrooms

The reality on the ground varies widely.

At the Beijing New Talent Academy, a private boarding school with fees of around $39,000 a year, students program humanoid robots and train on flight simulators. One 13-year-old there has used AI to build more than 200 apps. The school says AI helps him learn faster, and he tries to solve problems himself before asking for hints.

In rural Anhui province, a high school with 4,000 pupils, most of them children of migrant workers, uses AI for a different reason. It lacks qualified teachers, so students learn core subjects from AI-generated lessons on tablets. The headmaster says admissions to university rose from below 60% to 90%, but Reuters could not verify these figures. He estimates that at least 200 other Chinese schools are testing similar systems.

Observers describe the national picture as patchy. An executive at the education-technology company ClassIn told Reuters that few of the schools he has visited have woven AI into their teaching in a meaningful way.

.

Money Is Flowing, but Success Is Hard to Measure

Reuters reviewed more than 100 government procurement documents and found at least $31 million committed to AI classroom equipment, training and platforms in the past year. The real total is probably much higher, because spending is scattered across thousands of local budgets. The consultancy iResearch puts total 2025 spending on education digitalisation at just under $83 billion.

China has no standard way to measure AI literacy, so schools cannot easily tell whether their programs work. A researcher at the firm China Policy said that reliable assessment tools for school-level AI skills are still under development.

.

What the Research Shows So Far

The first studies point in different directions.

One study of nearly 27,000 Chinese teenagers linked AI use to an 18% rise in homework scores. The same students saw monthly exam scores fall by 20% within six months, with the steepest drops in social science subjects. A separate survey of more than 320,000 pupils found benefits, but also signs of dependence. Students who used AI at least three times a week scored lower on memory, comprehension, analysis and creativity than occasional users.

Neither study proves that AI caused the declines. Still, they feed a worry among teachers about "cognitive outsourcing," in which children hand mental work over to a machine instead of building the skill themselves. Chinese research reviewed by Reuters suggests that rural students may be especially at risk.

Some families are building their own safeguards. One Beijing mother lets her 11-year-old daughter try a hard maths problem for 15 minutes before turning to AI. A few days later, she asks the girl to solve it again without help.

.

Surveillance Under the Banner of Innovation

The Reuters review also points to a less discussed side of the project. Guizhou is bringing facial recognition and behavior analysis into classrooms, and Shanghai is introducing mental and physical health monitoring systems. In Yunnan, 14,000 classrooms are slated to have lessons recorded and analyzed by AI.

Such tools fit a pattern that predates the current AI push. In 2018, a Hangzhou high school drew international attention for using cameras that scored students' attentiveness and read their facial expressions. The South China Morning Post later reported that many Chinese schools and universities had adopted facial-recognition cameras and AI-powered gates under the education ministry's "smart campus" campaign. It also reported that this surveillance has raised privacy concerns among educators and parents.

China's education ministry did not respond to Reuters' request for comment on the issue.

.

Outlook

Beijing has set a clear target for 2030 but has not yet shown how it will measure progress. The next few years will show whether early gains in student engagement hold up or are outweighed by weaker exam results and over-reliance on AI.

Meanwhile, questions about student privacy are likely to grow louder as more classrooms add cameras and monitoring software. Other countries now face the same question of how quickly to bring AI into schools, and China's experiment will be watched closely for lessons, good or bad.

.

Sources

  1. Reuters, "Inside China's drive to build an AI-native generation" (Nov 16, 2026): https://www.reuters.com/world/china/inside-chinas-drive-build-an-ai-native-generation-2026-11-16/
  2. China Daily, "Guideline to regulate use of artificial intelligence in schools" (May 16, 2025): https://govt.chinadaily.com.cn/s/202505/16/WS6826de33498eec7e1f736de6/guideline-to-regulate-use-of-artificial-intelligence-in-schools.html
  3. Georgetown University CSET, English translation of the Ministry of Education's Guide to Using Generative AI in Primary and Secondary Schools (2025): https://cset.georgetown.edu/publication/china-school-gen-ai-use-guide-2025/
  4. Hangzhou municipal government portal, "AI takes root in basic education, nurturing future innovators" (Dec 25, 2025): https://www.ehangzhou.gov.cn/2025-12/25/c_296229.htm
  5. South China Morning Post, "Artificial intelligence is watching China's students, but how well can it really see?": https://www.scmp.com/news/china/politics/article/3027349/artificial-intelligence-watching-chinas-students-how-well-can
  6. South China Morning Post / Abacus, "High school uses facial recognition cameras to see if students are dozing off": https://amp.scmp.com/abacus/tech/article/3028514/high-school-uses-facial-recognition-cameras-see-if-students-are-dozing

.

What's Your Reaction?

Like Like 0
Dislike Dislike 0
Love Love 0
Funny Funny 0
Wow Wow 0
Sad Sad 0
Angry Angry 0

Comments (0)

User