S17.9 · Education, Training & Human Capital
A $67B out-of-school instruction market and the sector's starkest AI case, since the deliverable is a billable tutoring hour.
Tutoring and supplemental education covers out-of-school academic support and enrichment services, sized at $66.96B (2025, Fortune Business Insights), with APAC representing roughly 61% of global share on the strength of academic-competition demand in China and India. This is the starkest AI-disruption case in the sector, and the reason sits in the unit economics: the deliverable is, fundamentally, a billable human hour of instruction, and agentic AI is collapsing that cost line directly and quickly.
The growth profile is splitting down the middle. Historical growth ran double-digit in APAC and slower in the US and Europe; the forward picture bifurcates sharply between fast-growing AI-tutoring apps and flat-to-declining in-person tutoring — one category, opposite trajectories. Revenue concentrates in APAC — China, India and South Korea — while the US and Europe are smaller markets carrying a meaningfully higher price per hour. Fragmentation is the rule: branded chains such as Kumon, Sylvan and Varsity Tutors hold aggregate share alongside app-based players, but most tutoring hours are still delivered by independent tutors and local centers with no brand affiliation at all.
Curriculum content and tutor labor supply sit upstream; families paying directly sit downstream, alongside a growing channel of schools and districts contracting for supplemental services. Durability is short by sector standards — family retention through a single academic year or exam cycle, with no longer-duration subscription relationship — and margin is the billable-hour spread between what a tutor is paid and what the family is charged. Regulatory gating is comparatively light in most markets, typically background checks, but where it binds, it binds absolutely: China's 2021 ban on for-profit tutoring eliminated the domestic segment overnight. Light regulation in most markets does not make the segment regulation-proof everywhere.
EdTech platforms supply the delivery infrastructure; K-12 schools connect through contracted supplemental services; test preparation overlaps directly at the high-stakes-exam end of the market. School-district contracting is the defensive play, insulating tutoring providers against pure consumer-app displacement. AI-tutoring apps pushing into test prep and language learning is the growth play, carrying an existing AI-native product into adjacent subject areas.
Agentic AI collapses the tutor's billable hour toward zero more directly and more completely than almost anywhere else in the sector, because the deliverable is one-on-one instructional time rather than a licensed product or a regulated credential. The threat to human-delivery economics is head-on, not marginal. What holds is what an algorithm cannot supply: parent trust and human accountability specifically for younger children, existing school-contract relationships, and outcome guarantees that carry human oversight rather than purely algorithmic assurance.
Between the pure AI-tutoring app and the traditional human-staffed center, a new model is filling in — an AI tutor paired with human oversight, priced for parents who want AI-level cost and availability without giving up the human backstop. AI-tutoring substitution is already underway and accelerating; full displacement of human tutors in academic tutoring specifically is estimated at 2-5 years, faster than almost anywhere else in the sector, because comparatively little regulatory friction stands in the way.