Edtech in Asia 2026: After the Crackdown and Crash

Asian edtech in 2026 is a sector still picking itself up off the floor. The two defining events both came from this region: China banned its roughly $100 billion private-tutoring industry almost overnight in July 2021, and India’s BYJU’S — once Asia’s most valuable startup at $22 billion — collapsed into insolvency, declared “worth zero” by its own founder. Global edtech venture funding has fallen to about $2.4 billion in 2024, a decade low and roughly 89% below the 2021 peak (HolonIQ). And now generative AI is rewriting the rules again, turning free chatbots into the tutors the paid apps used to sell. This is an industry that boomed, broke, and is being rebuilt — and Asia is where all three happened first.

What happened to China’s edtech industry?

It was legislated out of existence in a single policy. On 24 July 2021, Beijing’s “double reduction” (双减) rules banned for-profit tutoring of core academic subjects for grades 1 to 9, forced existing companies to register as non-profits, and barred them from raising foreign capital or listing. The target was a sector worth roughly $100 billion. The day before the policy, the US-listed leaders cratered — TAL Education fell about 71% and New Oriental about 54% — and an industry employing more than 250,000 people at its top seven firms had its business model deleted (CNBC).

The survivors pivoted in ways nobody predicted. New Oriental, having lost roughly 90% of its market value, reinvented itself through East Buy — a livestream-commerce arm selling agricultural products with bilingual hosts, whose revenue jumped 651% to 4.5 billion yuan in its first full year. A tutoring giant became a shopping channel. Beijing has shown no sign of reversing the policy; as recently as 2025 it fined an unregistered operator to quash comeback rumours. Non-academic tutoring — STEM, coding, AI — has rebounded, but the old test-prep machine is gone for good.

Why did BYJU’S collapse?

India’s BYJU’S is the cautionary tale the whole sector now studies. It peaked at a $22 billion valuation in 2022, the most valuable startup in India and a symbol of edtech’s pandemic boom. Then it unravelled: BlackRock cut its valuation by roughly 95% to about $1 billion in January 2024, its founder admitted in October 2024 that it was “worth zero,” and it fell into insolvency over an unpaid debt to the national cricket board, compounded by a $1.2 billion dispute with US lenders. As of 2026 it remains in insolvency with no resolution plan, its last audited accounts dating to 2022.

BYJU’S didn’t fall alone, but the survivors prove the model can still work. PhysicsWallah — built on cheap, high-volume online coaching for India’s exam-obsessed students — IPO’d in November 2025, closing its first day up around 44% at a roughly $5 billion valuation. Indian edtech funding tells the boom-bust arc in one line: from a 2021 peak near $4.6 billion to just $283 million in 2023, an 88% drop (Inc42). The capital left; the demand didn’t.

Who is winning in Indian edtech now?

The survivors share a formula: cheap, high-volume, exam-focused, and built for a market that genuinely needs it. India produces millions of students sitting brutally competitive entrance exams every year, and the durable edtech businesses are the ones serving that demand at a price point the mass market can afford — not the premium, content-library model BYJU’S over-built.

PhysicsWallah is the template. Founded by a teacher who started on YouTube, it undercut the incumbents on price, went profitable, and IPO’d in November 2025 at roughly $5 billion — a vindication of lean economics over blitzscaling. Unacademy, once a $3.4 billion-valued rival, has pulled back hard, cutting losses and reportedly exploring a sale of parts of the business — the other side of the correction. upGrad in the higher-ed and upskilling segment, and Vedantu and Allen in test prep, round out a market that is smaller, more sober and more profitable than the 2021 frenzy. The lesson Indian edtech learned the hard way: the unit economics have to work before the valuation does.

How big is the Asian edtech market now?

Smaller and humbler than the forecasts promised. The headline figures from the boom years — HolonIQ’s much-cited “$404 billion global edtech spend by 2025” — were 2020-era projections that no longer describe reality, and should be read as historical optimism rather than current fact. Research-firm estimates put the Asia-Pacific edtech market around $36.5 billion in 2024, growing toward $88 billion by 2030 (Grand View Research), though these vendor numbers use broad, inconsistent definitions and are best treated as directional.

The Southeast Asian scene has thinned dramatically. Indonesia’s Zenius — one of the region’s oldest edtech firms — shut down entirely in January 2024 after more than two decades, having already gone through multiple rounds of layoffs. Ruangguru, the Indonesian leader still operating, cut hundreds of jobs and faded from its 2021-vintage near-billion-dollar valuation. The pattern repeats across the region: pandemic-era funding inflated a wave of tutoring and test-prep startups, demand normalised when schools reopened, and the capital left faster than it arrived.

Vietnam is the most active second-tier market, with strong, durable demand for English-language learning and exam prep — ELSA, the AI pronunciation-coaching app, and a wave of online English platforms serve a young population that treats English fluency as an economic ticket. Singapore, with its small but affluent and education-obsessed market, hosts edtech startups and the regional headquarters of larger players, and the Philippines remains the world’s hub for online English tutoring, supplying teachers to learners across Asia. But no breakout regional champion has emerged from Southeast Asia the way Grab did in mobility or Sea did in gaming. Tellingly, the Google-Temasek-Bain e-Conomy report — the definitive scorecard of Southeast Asia’s digital economy — stopped tracking edtech as a standalone sector after 2021. The demand is real and structural; what’s missing is a scaled, profitable champion to capture it.

What about Japan and Korea?

The two wealthiest North Asian education markets are where edtech is most mature — and where the AI-in-classroom experiment has gone furthest. South Korea ran one of the boldest policy moves in global edtech: a national rollout of AI-powered digital textbooks in schools from 2025, the first country to attempt it at that scale — before parental and teacher backlash over screen time and data forced the government to walk it back to optional “educational material” status within a year. It’s the mirror image of China’s tutoring ban: where Beijing legislated edtech out, Seoul tried to legislate it in, and both ran into the limits of policy-driven education change. Korea’s private scene is strong too, led by Mathpresso’s QANDA, a math-solving AI app with global reach.

Japan is a steadier, more institutional market — its edtech sector was worth roughly $14.8 billion in 2024 and is dominated by established names: Benesse (the textbook and tutoring giant), Recruit’s Quipper, and adaptive-learning startup Atama Plus. Japan’s edtech reflects its broader tech character — methodical, enterprise-and-school-led, less prone to the boom-and-bust venture cycles that whipsawed India and Southeast Asia. Between Korea’s policy experiments and Japan’s institutional depth, North Asia is where edtech is being absorbed into the school system rather than sold around it.

How is AI changing edtech in Asia?

This is the disruption the sector is still absorbing, and it cuts both ways. The threat is existential for paid tutoring: when a free chatbot can explain calculus or check an essay, the value of a subscription erodes. The share of US teens using ChatGPT for schoolwork doubled to 26% in 2024, and the global cautionary tale — Chegg — saw its stock fall 48% in a single day in 2023 after blaming ChatGPT for its decline (Pew; Fortune). Every Asian test-prep app faces the same squeeze.

The opportunity is rebuilding tutoring around AI rather than against it. China’s TAL pivoted into AI learning hardware; ByteDance’s Gauth became one of the top education apps in the US; India’s PhysicsWallah launched its own AI tutor. The likely winners aren’t the apps that sold content — they’re the ones that turn AI into a personalised, always-available teacher at a price the mass market can pay. For the broader regional AI picture, see our Asia AI market breakdown and how each market accesses AI tools.

Frequently asked questions

Why did China ban private tutoring?

In July 2021, China’s “double reduction” policy banned for-profit tutoring of core academic subjects for grades 1 to 9, aiming to reduce student workload and family education costs and ease inequality. It forced tutoring firms to become non-profits and barred them from foreign capital, wiping out a roughly $100 billion industry almost overnight.

What happened to BYJU’S?

BYJU’S, once valued at $22 billion as India’s most valuable startup, collapsed — BlackRock cut its valuation roughly 95% to about $1 billion in January 2024, its founder called it “worth zero,” and it fell into insolvency. It remains in insolvency proceedings in 2026.

How big is Asia’s edtech market?

Research-firm estimates put the Asia-Pacific edtech market around $36.5 billion in 2024, growing toward $88 billion by 2030 (Grand View Research), though these are vendor projections with broad definitions. Global edtech venture funding fell to a decade-low ~$2.4 billion in 2024.

Is AI a threat to edtech companies?

Both. Free AI chatbots erode the value of paid tutoring — the share of US teens using ChatGPT for schoolwork doubled to 26% in 2024, and Chegg’s stock fell 48% in a day after blaming ChatGPT. But edtech firms are also rebuilding around AI tutors, which may be the sector’s path back to growth.

Which Asian edtech companies are winning in 2026?

The survivors are lean and exam-focused: India’s PhysicsWallah IPO’d at ~$5 billion in November 2025, China’s New Oriental rebuilt through livestream commerce, and non-academic tutoring (STEM, coding, AI) has rebounded in China where core-subject tutoring was banned.

—