In the Age of AI, China Is Rethinking English and What Foreign-Language Education Is Really For

Close-up of a person writing notes from an open book, emphasizing education and studying.

For generations of Chinese students, English has been far more than another school subject.

It has been a gateway to universities, international business, scientific research, overseas travel and a changing global economy. Since China’s opening to the outside world in the late 1970s, English has occupied a prominent position in education and has become one of the country’s core academic subjects.

Now artificial intelligence is forcing a difficult question into the classroom:

If AI can translate almost instantly, is learning English still worth the enormous amount of time Chinese students spend studying it?

That question became a national online debate in September 2026 after Tang Jiafeng, a Chinese math teacher and education blogger, called for English to become optional in schools. He argued that students can spend years studying the language before becoming genuinely proficient, while AI translation tools increasingly allow people to communicate across language barriers without mastering another language.

His argument quickly collided with a very different view.

For supporters of continued English education, the language remains an essential skill—not because Chinese people need to imitate Western culture, but because English is still deeply embedded in global science, technology, higher education, business and information exchange.

That disagreement has turned a school-curriculum question into a much larger debate about China’s future.

Why English Became So Important in China

China’s enthusiasm for English did not emerge by accident.

Following the country’s opening and reform period beginning in the late 1970s, learning English became increasingly associated with modernization and international engagement.

English was eventually incorporated into the national education system as one of the three major academic subjects alongside Chinese and mathematics. It also became an important component of the gaokao, China’s intensely competitive national college entrance examination.

For many families, English proficiency represented opportunity.

A student who knew English could potentially access foreign research.

They could apply to overseas universities.

They could communicate with international companies.

They could read scientific literature without waiting for translation.

And they could participate in an increasingly globalized economy.

For several decades, therefore, English was not viewed simply as a language.

It was an investment.

But the Economic Calculation Is Changing

That investment is now being reconsidered.

Chinese students face intense competition for university places and increasingly uncertain employment prospects after graduation. At the same time, AI systems are becoming capable of translating documents, conversations and online content with unprecedented speed.

That creates a new economic calculation.

If a student spends hundreds or thousands of hours mastering English, what is the return on that investment?

Would those hours be better spent learning mathematics, science, programming, engineering, AI or another practical skill?

The question becomes particularly uncomfortable when students are studying English primarily to pass exams rather than to communicate.

This is one reason the debate has resonated so strongly online.

It is not really a debate about vocabulary lists.

It is about opportunity cost.

China’s English Problem May Not Actually Be English

Some critics of the current system argue that China has spent too much effort teaching English as a test subject instead of a communication skill.

The New York Times quoted education researcher Xiong Bingqi, who described the result as “dumb English”—years of instruction that may produce examination skills without comparable real-world proficiency.

That criticism is important.

A student can memorize thousands of English words and still struggle to have a conversation.

They can score well on grammar questions while finding it difficult to write naturally.

They can understand reading passages yet hesitate when asked a simple question by a foreigner.

This is not unique to China.

Many countries struggle with the difference between language education and language proficiency.

But China’s enormous exam culture makes the problem particularly visible.

AI Translation Attacks the Original Justification

Artificial intelligence is changing this equation in a way that previous technologies did not.

Computers have translated text for years.

Online translation services have existed for decades.

Smartphone translation became increasingly convenient in the 2010s.

But modern generative AI can now process context, rewrite phrases, interpret conversations and generate responses rather than simply converting words one by one.

That makes the technology feel qualitatively different.

A traveler can point a smartphone at a sign.

A businessperson can translate an email.

A conference participant can receive captions in another language.

A student can ask AI to explain an English sentence.

A company can translate large amounts of content almost instantly.

And real-time speech translation is improving quickly.

This is why the old argument—”you must know English because translation is difficult”—is becoming less convincing.

China Is Already Integrating AI Into Education

There is a larger irony here.

China is not rejecting AI in education.

Quite the opposite.

In April 2026, China’s Ministry of Education and four other government agencies issued an “AI + Education” action plan calling for AI to become deeply integrated into teaching, learning, research and educational management by 2030. The plan envisions more personalized learning, stronger AI literacy and new education models built around human-machine collaboration.

The Ministry of Education also identified AI integration as a major 2026 priority, emphasizing personalized learning and interdisciplinary skills.

This means the English debate is taking place inside a much bigger educational transformation.

China is already asking schools to prepare students for an AI-driven economy.

The difficult question is deciding which traditional skills should remain foundational.

English is simply one of the most visible examples.

AI Can Translate Language, but Can It Replace Language Ability?

This is the heart of the argument.

Suppose an AI system can translate a conversation between Chinese and English.

Does that make English proficiency unnecessary?

Not necessarily.

Translation transfers language.

Language knowledge transfers understanding.

Those are related but different capabilities.

Someone who knows English can directly understand a paper, meeting or conversation.

Someone relying entirely on AI needs an intermediary.

That distinction matters when accuracy, speed or nuance is important.

It also matters when AI is unavailable.

And it matters when the person needs to judge whether the translation itself is correct.

The Translator Still Needs to Know the Language

This becomes especially important when dealing with AI-generated translations.

A person who does not understand English may have difficulty recognizing a mistranslation.

That is a fundamental problem.

Imagine an AI translates a business contract incorrectly.

If you know the source language, you can notice that something sounds wrong.

If you do not, you may have no way to know.

Researchers studying Chinese university students’ use of AI in translation have found that AI tools can support translation learning, but they also raise questions about learner engagement, competence and how humans should work with AI rather than simply handing tasks over to it.

Another 2026 study of Chinese university translation students found that a structured AI-collaborative workflow improved translation performance and learner engagement compared with conventional instruction. The finding suggests an important middle path: AI can change how languages are learned without necessarily making language knowledge obsolete.

English Is Still the Language of a Huge Global Knowledge System

There is another reason English remains valuable.

Even if AI can translate information, the global information ecosystem itself still contains enormous amounts of English-language material.

Scientific research.

Technical documentation.

Programming resources.

International conferences.

Academic journals.

Corporate communications.

Diplomatic material.

Open-source projects.

Global news.

Professional communities.

A researcher who speaks English can interact directly with this ecosystem.

An AI translation layer can help, but it remains an additional step.

This is particularly relevant in science and technology.

English proficiency can allow a researcher to read an emerging paper immediately rather than waiting for translation or relying on a summary.

That advantage may become smaller as AI improves.

But it has not disappeared.

AI Could Actually Increase the Value of English for Some People

This seems counterintuitive.

If translation becomes easier, why would English become more valuable?

Because the ability to use AI effectively may itself depend on language.

Researchers and English teachers cited by The New York Times argued that some leading AI models and resources are more accessible or effective in English.

The quality gap may narrow as Chinese-language models improve.

Chinese AI companies have made rapid progress in large language models, translation and multimodal systems.

But international AI research is still heavily multilingual, with English remaining an important bridge language.

That creates a paradox:

AI can reduce the need to speak English while simultaneously creating new reasons to understand English.

The Internet Has a Similar Problem

Consider how the internet evolved.

Automatic translation makes foreign websites easier to read.

But English remains one of the most important languages online.

A person who knows English can search directly.

They can compare sources.

They can participate in global communities.

They can communicate with international users.

They do not need to wait for translation.

AI reduces the friction.

It does not necessarily eliminate the value of direct access.

What About Travel?

Travel is another test case.

Translation tools are already transforming tourism.

A traveler can translate menus.

They can use speech-to-speech translation.

They can ask an AI assistant to interpret signs.

They can communicate basic needs without knowing the local language.

That is a major improvement.

But travel also demonstrates why language knowledge remains useful.

Knowing basic English can help a Chinese traveler navigate airports, hotels and international tourist destinations without depending completely on technology.

It can also create spontaneous conversations.

Technology can translate a sentence.

It cannot always reproduce the social confidence that comes from understanding the language directly.

AI Translation Is Not Perfect

The debate would be much simpler if AI translation were flawless.

It is not.

Current AI translation systems can struggle with ambiguity, specialized terminology, dialects, slang, humor, cultural references and low-resource languages.

They can also produce fluent-sounding errors.

That last problem is particularly dangerous.

An awkward machine translation is easy to distrust.

A polished but incorrect translation can be much harder to detect.

For high-stakes areas such as law, medicine, engineering and finance, human review remains important.

This does not mean AI translation is unreliable in general.

It means reliability depends heavily on the context.

The Hidden Value of Learning a Language Is Not Translation

Language education also develops abilities that machine translation cannot simply replace.

Learning another language exposes students to different ways of organizing ideas.

It creates familiarity with another culture.

It can improve awareness of communication patterns.

It teaches students to tolerate ambiguity.

It requires sustained memory and practice.

And it provides direct access to another community.

These are educational benefits, not merely translation benefits.

A student studying English is not only acquiring a tool for converting Chinese into English.

They are learning how English-speaking people communicate.

That can matter in diplomacy, international business, research and cultural exchange.

The National Identity Dimension

The argument has become politically charged in China partly because language carries symbolic meaning.

Tang Jiafeng’s suggestion that Chinese students were engaged in “blind worship” of foreign imports struck a sensitive nerve, according to The New York Times, because that language can imply excessive deference to foreign powers. Former Global Times editor Hu Xijin pushed back against what he described as an overly nationalist interpretation of the issue.

This illustrates that English in China has always had two identities.

It is a practical tool.

And it can be a cultural symbol.

For some people, emphasizing English represents China’s openness to the world.

For others, reducing its dominance represents greater confidence in China’s own language and culture.

The AI debate brings that tension back to the surface.

China’s Changing Relationship With the West Adds Another Layer

There is also a geopolitical dimension.

The world of the late 1970s was dramatically different from the world of 2026.

China has become a global economic and technological power.

Its universities produce significant amounts of research.

Its technology companies compete internationally.

Its domestic market is enormous.

Chinese technology platforms and AI models are increasingly influential.

The old assumption that modernization necessarily required adopting Western practices is therefore being questioned.

That does not mean China is becoming isolated.

It means the rationale for learning English has changed.

The country no longer needs English primarily as a symbol of catching up.

The question is whether English remains useful as a tool for participating in global systems.

Shanghai Already Reduced Some English Pressure

The national debate is not starting from zero.

China has already experimented with reducing the pressure surrounding English.

In 2021, Shanghai stopped using English as a final examination subject for elementary school students as part of efforts to reduce academic pressure. A government adviser also proposed removing English from the gaokao, although that proposal did not become national policy.

These moves reflect a broader educational concern:

Should children spend so much time preparing for tests?

The same question has been applied to mathematics, science and other subjects.

English is therefore part of a much wider debate about China’s highly competitive education system.

Some Universities Are Going Further

The shift is becoming more visible in higher education.

In July 2026, a research university in Shenzhen announced that English-language courses would be phased out and replaced with courses comparing Chinese and Western cultures, explicitly citing AI’s growing translation capabilities.

That is a much more radical step than reducing elementary-school English exams.

It suggests that some educators are beginning to ask a fundamental question:

Should universities teach English as a general-purpose skill when AI can increasingly provide translation on demand?

The answer is likely to vary by discipline.

An engineering student may need different language abilities from a literature student.

A computer scientist working with international research may need more English than someone whose career is entirely domestic.

A tourism professional may need conversational English.

A government employee dealing with international affairs may require advanced fluency.

This points toward a possible future of more targeted language education.

The Future May Be “English for Purpose”

Instead of treating English proficiency as a universal requirement, education systems could increasingly tailor it to students’ needs.

Students likely to pursue international science could receive advanced English training.

Students entering international business could focus on negotiation, presentations and professional vocabulary.

Tourism students could learn practical spoken communication.

Researchers could learn how to read academic papers.

Students whose future careers are entirely domestic might receive a lighter language curriculum.

AI could provide the support layer for everyone else.

This would preserve English where it has high economic value while reducing unnecessary test-driven instruction.

But That Could Create an Inequality Problem

There is a serious argument against simply removing English from public education.

English education is expensive when families have to buy it privately.

Wealthier families can hire tutors.

They can enroll children in international schools.

They can arrange overseas experiences.

They can provide access to English-speaking environments.

Poorer families may depend almost entirely on public schools.

This is exactly the concern raised by English teacher Cinderella Huang in the New York Times report. She argued that removing English as a core subject could widen the gap because affluent families would continue buying language education while poorer students would lose their main opportunity to learn it.

That is one of the strongest arguments for retaining broad access to English even if the teaching method changes.

A skill can become less essential for everyone while remaining highly valuable to a smaller group.

If public schools stop teaching it, access may become more unequal.

AI Could Make the Inequality Problem Better—or Worse

AI translation is relatively easy for people with smartphones and internet access to use.

That could democratize communication.

A student from a rural area might use AI to read an English research paper.

A small business could communicate with foreign customers.

A traveler could navigate another country.

That is the optimistic scenario.

The pessimistic one is that wealthier students learn both English and AI while poorer students rely entirely on automated translation.

In that case, technology may actually increase the advantage of direct language proficiency.

The future therefore depends partly on who has access to good AI tools and who learns how to use them.

China’s AI Education Strategy May Change the Argument

The government’s 2026 AI + Education Action Plan offers an important clue.

China is not treating AI as a replacement for basic human knowledge.

The plan emphasizes human-centered education, AI literacy, interdisciplinary skills and human-machine collaboration. Its stated goal is to develop highly capable people for the intelligent era while combining knowledge transmission with ability development and humanistic education.

That philosophy could apply directly to language.

Rather than asking whether AI or English wins, schools could teach students to combine both.

Students learn enough English to understand, evaluate and communicate.

AI handles repetitive translation.

Humans handle judgment, context and relationships.

That model may be more realistic than either extreme.

The Question Is Not “Can AI Translate English?”

It obviously can.

The more useful question is:

What becomes possible when a person knows English and also has AI translation?

That combination could be far more powerful than either skill alone.

A bilingual scientist can use AI to explore additional languages.

A bilingual entrepreneur can negotiate more efficiently.

A bilingual traveler can use AI to communicate in places where English does not work.

A bilingual student can verify AI-generated translations rather than blindly trusting them.

In this model, AI does not eliminate language skills.

It amplifies them.

What Happens to English Exams?

This may be where the biggest reform eventually occurs.

If education authorities decide that traditional grammar-heavy testing no longer reflects real-world value, English could be assessed differently.

Instead of memorizing isolated vocabulary, exams could measure:

Reading comprehension.

Listening.

Speaking.

Writing.

Cross-cultural communication.

Information verification.

AI-assisted communication.

The ability to identify translation errors.

That would produce a very different kind of English education.

Students would still learn the language.

But they would learn it for use rather than for test performance.

English Might Become Less Important as a Subject and More Important as a Tool

This sounds contradictory, but it is probably the central direction of change.

English could become less dominant as a universal academic hurdle while remaining highly useful professionally.

That would mirror what happened with mathematics and computer skills in other contexts.

Not everyone becomes a mathematician.

But basic mathematics remains essential.

Not everyone becomes a programmer.

But digital literacy is increasingly valuable.

English could follow a similar pattern.

Universal testing may decline.

Targeted practical proficiency may become more important.

The Same Debate Could Spread Around the World

China is not alone in asking whether AI makes language learning less necessary.

As speech translation becomes cheaper and more accurate, companies, universities and travelers everywhere will confront the same question.

Do people really need to learn another language if their phones can translate conversations?

The answer will vary depending on the goal.

For ordering dinner, perhaps not.

For reading literature, negotiating a complex deal, conducting research, building relationships or living permanently in another society, language knowledge remains more consequential.

AI will reduce the cost of communication.

It will not make communication identical to understanding.

The Human Relationship Problem

There is another reason language may survive AI translation: relationships.

Imagine meeting someone from another country.

You could communicate entirely through an AI translator.

But the experience of speaking even imperfectly in another person’s language can signal effort, respect and curiosity.

People often respond differently when someone has taken the trouble to learn their language.

That is difficult to automate.

A machine can transmit meaning.

It cannot fully reproduce the human meaning of learning someone’s language.

English Could Become a Voluntary Investment Rather Than a Mandatory Burden

This may ultimately be the direction China takes.

Instead of asking every student to reach the same level of English proficiency, the country could provide a universal foundation and allow students to specialize based on their future goals.

Those headed toward international careers could go much further.

Others could focus on Chinese-language mastery, mathematics, science, technology, arts or vocational skills.

AI would serve as a translation safety net.

Such a system could reduce educational pressure without completely abandoning international connectivity.

Why China Should Be Careful About Moving Too Quickly

There is still a risk in assuming technology will solve every language barrier.

AI systems can change rapidly.

Policies can change quickly.

But education takes years to develop.

A student who stops learning English today may enter a global workplace a decade from now under conditions nobody can predict.

Perhaps AI translation will be nearly flawless.

Perhaps English will remain indispensable in science and technology.

Perhaps both will be true simultaneously.

Education policy therefore has to account for uncertainty.

Removing a foundational skill is difficult to reverse.

The Real Lesson May Be About How Education Should Adapt to AI

The English debate is ultimately a test case for a much larger educational problem.

AI is making some forms of knowledge cheaper.

It can translate.

Summarize.

Calculate.

Generate text.

Write code.

Explain concepts.

Search information.

Create images.

As that happens, schools have to reconsider what humans should spend their limited learning time mastering.

Should students memorize more information?

Or should they learn how to evaluate information?

Should they practice repetitive tasks?

Or learn how to supervise AI?

Should they focus on producing answers?

Or on asking better questions?

English sits directly inside this transformation because AI can perform a task that once required years of human training.

The technology therefore forces educators to separate the parts of language learning that machines can automate from the parts that humans still need.

The Most Important Skill May Be Knowing When Not to Trust AI

Ironically, that could strengthen the case for education.

The more powerful translation becomes, the more valuable verification becomes.

If an AI system translates ten thousand words in seconds, humans still need to determine whether the translation is correct.

That requires knowledge.

A person who understands both languages is in a stronger position to judge the output.

The future may therefore belong not to people who refuse AI and insist on doing everything manually, but to people who understand enough to supervise AI effectively.

China’s English Debate Is Really a Debate About the Future

The argument over English is not going away.

Some parents want less testing.

Some teachers want more practical language education.

Some policymakers see opportunities to reduce student workloads.

Some educators worry about international isolation.

Some families fear greater inequality.

And AI developers continue to make translation cheaper and more capable.

All of these concerns can be true at once.

The most interesting outcome may therefore not be the disappearance of English from Chinese schools.

It may be the end of the assumption that every student must study English in exactly the same way for exactly the same reason.

China is entering an era in which AI can increasingly perform the mechanical part of language conversion.

That allows education to focus more deliberately on what remains uniquely human:

Understanding context.

Evaluating meaning.

Building relationships.

Thinking critically.

Understanding other cultures.

And communicating with intention.

The Future of English in China May Be Smaller—but Smarter

English once represented China’s opening to the world.

In the AI era, it may represent something different: a strategic capability rather than a universal obligation.

That could mean fewer hours spent memorizing grammar for an exam and more emphasis on practical communication.

It could mean English becoming more specialized.

It could mean AI translation becoming a routine classroom tool.

It could mean students learning how to compare machine translations and identify mistakes.

It could mean more attention to Chinese-language mastery alongside foreign-language capability.

And it could mean that the value of English is increasingly determined by what a student wants to do rather than by a single national examination.

The keyboard did not disappear when speech recognition arrived.

Calculators did not make mathematics worthless.

Search engines did not eliminate the value of reading.

Likewise, AI translation may not make language learning obsolete.

Instead, it may force the education system to answer a better question:

What parts of learning a language are worth doing yourself when a machine can increasingly do the translating for you?

That is a question China is now asking.

And eventually, almost every education system in the world may have to answer it.

5 Frequently Asked Questions

1. Why are people in China questioning the value of learning English?

AI translation has made cross-language communication much easier, while Chinese students are also facing heavy academic workloads and uncertain employment prospects. A viral proposal to make English optional in schools reignited an older debate over whether English takes too much classroom time relative to its practical value for students who will work primarily in China.

2. Is China planning to stop teaching English in schools?

There is no announced nationwide policy to eliminate English education. China has already reduced some English-testing pressure in certain contexts, however, and individual institutions have experimented with reducing or replacing English courses. In July 2026, a Shenzhen research university announced plans to phase out English courses in favor of courses comparing Chinese and Western cultures, citing improvements in AI translation.

3. Can AI translation replace learning English?

For many everyday situations, AI can significantly reduce the need to know English. But translation does not completely replace language proficiency. Direct English knowledge remains useful for research, higher education, international business, travel, cultural understanding and verifying whether AI-generated translations are accurate.

4. Could reducing English education increase inequality?

Yes, critics have raised that possibility. Wealthier families can continue paying for tutors or international schools if English becomes less important in public education. Students from lower-income families may then have fewer opportunities to acquire English proficiency. The New York Times reported concerns that public-school English may serve as an important equalizing opportunity for children who cannot access private language education.

5. What could English education look like in China’s AI era?

One possible direction is a shift from exam-heavy universal instruction toward more practical and targeted language education. Students pursuing international careers could receive advanced English training, while others could rely more heavily on AI translation. Schools could also teach students how to use AI translation tools, evaluate machine-generated language and understand cultural context. This approach would fit China’s broader 2026 “AI + Education” strategy, which emphasizes human-AI collaboration, personalized learning and interdisciplinary skills.

Sources The New York Times

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top