When China decided to take on its own tutoring industry in mid-2021, the move looked, from across the border, like a thunderclap.
For someone working in India’s education space, it was difficult not to watch with disbelief.
Overnight, Beijing turned a thriving, investor-driven, ultra-competitive industry into a public utility.
The same billion-dollar companies that had built after-school empires were ordered to become non-profits, stop teaching school subjects, and never again raise foreign capital.
In India, we often talk about “regulation catching up with innovation.”
In China, regulation sprinted ahead and knocked innovation down.
Four years later, in 2025, that fallen empire has quietly rebuilt itself — not through rebellion, but through reinvention.
2021: The Day the Apps Went Dark
I still remember a Shanghai-based colleague sending me a photo of a tutoring street — entire rows of coaching centres suddenly shuttered, their bright red banners rolled up like political posters after an election.
The government’s “Double Reduction” policy (双减政策), announced in July 2021, aimed to cut homework and after-school tutoring for children in grades 1 to 9.
The idea, officials said, was to reduce stress on families and lower education costs — both seen as key reasons behind China’s falling birth rate.
The result was brutal.
Companies like TAL Education and New Oriental, once valued in tens of billions of dollars, saw their stocks crash by over 90%.
Tens of thousands of tutors lost jobs.
Parents who had built their routines around evening classes were suddenly left without options.
In one Beijing district, the police even posted notices warning parents against hiring “illegal private tutors.”
It wasn’t just an economic shift.
It was a cultural shock: China had outlawed something as basic as studying more.
2022–2023: The Gray Years
The first year after the crackdown was quiet on the surface, but full of underground activity.
Through education circles, you heard stories of parents forming micro-learning groups in apartment complexes.
Retired teachers began taking lessons on WeChat audio calls.
Some tutoring apps rebranded themselves as “family education consultants.”
A few found creative loopholes — teaching in English, or through “coding logic” classes that just happened to mirror the math syllabus.
By 2023, another pattern emerged: hardware.
Instead of coaching apps, the market pivoted to learning devices — AI pens, adaptive-learning tablets, voice-activated dictionaries.
When you can’t teach a student directly, you can always sell them a tool that “helps them learn independently.”
And that, conveniently, fit the government’s push for AI industrialisation.
Companies like iFlyTek and TAL’s new hardware division started marketing learning gadgets for “smart family learning.”
It wasn’t tutoring — at least not officially.
But functionally, it was.
Parents I spoke with during an education fair in Guangzhou last year admitted that their children now “study through devices” almost every evening.
The teacher is gone; the algorithm has replaced her.
2024: The Quiet Comeback
By 2024, the mood had changed.
Officially, the rules hadn’t softened — the 2021 policy still stood.
But the enforcement had relaxed.
Local authorities began quietly allowing licensed centres to reopen under new categories like “enrichment training” or “talent development.”
New Oriental, once the poster child of the tutoring boom, reinvented itself as a live-stream commerce company — selling English courses alongside cooking oil and rice cookers.
Its founder, Yu Minhong, turned into a kind of national folk hero for resilience.
The company’s education spin-off, Koolearn, went public again, now rebranded as an “AI learning service provider.”
Meanwhile, ByteDance, which had disbanded its education arm in 2021, quietly integrated AI tutoring features into its productivity app Feishu.
No one called it “after-school tutoring.”
It was just “AI-assisted note review.”
The line between compliance and creativity had blurred.
The same state that once dismantled EdTech now saw it as a crucial player in China’s AI race — as long as it didn’t challenge public education.
The Human Side of Reinvention
In Chengdu, a former math tutor I spoke to over WeChat voice chat described her transformation in simple terms:
“Before, I taught children to solve equations. Now, I teach AI to explain equations.”
She works for a startup that trains voice models for educational chatbots — a job that didn’t exist in 2021.
She earns slightly less than before, but she doesn’t fear being “raided” for teaching the wrong subject.
Students, too, have adapted.
A 13-year-old boy in Shenzhen showed me his study setup on a video call — a desk, a smart lamp that tracks his posture, a tablet that reads his expressions to gauge attention, and a digital pen that marks his errors in real time.
“I don’t have a teacher anymore,” he said cheerfully, “but my pen scolds me if I get lazy.”
It’s a strange hybrid world — part surveillance, part self-learning, and all quietly tolerated.
2025: What the Numbers Say
China’s Ministry of Industry and Information Technology now classifies “smart education” as a strategic growth sector.
According to iResearch China, the country’s EdTech hardware market is projected to reach ¥79 billion (around USD 11 billion) by 2028, up from ¥43 billion in 2023.
More than 85% of traditional online tutoring firms, however, remain closed.
The new EdTech 2.0 operates with a government-approved vocabulary: “AI literacy,” “intelligent learning,” “vocational reskilling.”
The term “tutoring” has disappeared — but the behaviour it described never did.
Did the Crackdown Work?
Four years later, it’s hard to say whether the “Double Reduction” policy achieved its goals.
It did reduce the number of formal tutoring hours and probably made after-school evenings calmer for many families.
But it didn’t really reduce pressure — it just redistributed it.
Parents still compete, now through devices, test-prep books, and paid AI subscriptions.
The social energy of competition — to get ahead, to out-perform — didn’t vanish.
It only found new, more discreet channels.
One Beijing parent told me:
“The policy said no tutoring. Fine. But if my child doesn’t prepare, others will — with AI apps or through ‘enrichment camps.’ So what choice do we have?”
It’s a sentiment that sounds familiar in India, where exam pressure and coaching culture run deep.
The difference is, China tried to legislate its way out of it — and found that human ambition is harder to regulate than corporate structure.
Lessons for India
For India, watching this from the sidelines has been instructive.
Our own EdTech market — sprawling, diverse, and largely unregulated — could easily swing between over-commercialisation and over-correction.
If China’s story says anything, it’s that governments can change the rules overnight, but education ecosystems evolve slowly, adapting through loopholes, language, and technology.
India’s current push for AI-enabled classrooms and digital public infrastructure must walk a fine line: encourage innovation, but not inequality.
Because if the Chinese experience shows one thing, it’s this —
When you suppress one form of learning, another emerges, often smarter, subtler, and harder to control.
Epilogue
Walking past a bookstore in Beijing this year, I noticed a new kind of shelf: not test-prep, not textbooks — but AI study companions.
Each device claimed to “personalise learning with government-approved ethics.”
That phrase stuck with me.
The 2021 crackdown tried to tame the chaos of private tutoring.
It succeeded in changing its shape, but not its spirit.
Four years on, China’s classrooms may look quieter — yet somewhere in the glow of every smart tablet, the same old race continues, only this time, the invigilator is an algorithm.