S17.3 · Education, Training & Human Capital
The sector's horizontal software infrastructure, sized around $187B, where AI collapses content-authoring cost but procurement relationships hold.
EdTech and digital learning platforms are the software layer delivering instruction and course content across K-12, higher education, corporate and consumer markets. Grand View Research sizes the segment at $187.0B (2025); other houses land anywhere from $130B to $250B, a spread reflecting inconsistent scoping of K-12, higher-ed and corporate learning rather than genuine disagreement about underlying demand. This is the sector's horizontal infrastructure — the platform buyers expand from into every adjacent content vertical — and the tension defining it is simple: agentic AI is compressing content-authoring cost toward zero while procurement relationships do the work of holding the category together.
Post-COVID growth ran ~13-16% CAGR and is normalizing to a still-strong ~10.8% through 2033, reaching an estimated $437.5B, on AI-native platforms and rising device penetration in emerging markets. Revenue concentrates in the US and China; the fastest growth is APAC — India and Southeast Asia in particular — and it is mobile-first by default, not by strategic choice. Concentration splits by layer. Learning management and student information systems — PowerSchool, Instructure, Coursera, Duolingo — form a moderately concentrated platform tier carrying meaningfully higher lock-in; the content-app layer beneath remains a long, fragmented tail.
Content publishers and cloud infrastructure providers sit upstream; schools, districts, universities, individual consumers and employers sit downstream, with the employer channel crossing straight into corporate learning demand. Platform-layer economics are genuinely SaaS — 60-80% gross margins, low capital intensity — and durability is net revenue retention and seat renewal in the literal subscription sense, not by analogy. The gates are student-data privacy regimes (FERPA, COPPA), multi-year district procurement cycles and accessibility compliance (ADA/WCAG). All three slow a new entrant's sales cycle more than they slow anyone's product development, which makes them a moat for whoever already holds the contract.
Adjacency here is positional: EdTech touches every vertical content segment in the sector — K-12, higher education, corporate learning, tutoring, language learning — and platform vendors work that position two ways. LMS and SIS incumbents add vertical capability to raise the cost of leaving the system of record, defending revenue they already bill. Content vendors add subjects and credentials to grow average revenue per user without adding a single new institutional customer.
Agentic AI takes content-authoring and courseware-localization cost to near zero — work that once required instructional designers and subject-matter writers can now be substantially automated. That erodes the content-library moat many platforms were built on: a large existing library of courses is worth less when a competitor can generate a comparable one on demand. What survives is the relationship and the compliance record — procurement relationships built over multi-year sales cycles, data-privacy compliance infrastructure, and credibility on proctoring and assessment integrity. None of it can be manufactured quickly, and all of it renews.
The forming category sits past content delivery: AI tutoring agents behaving less like a platform and more like an instructor — interactive, adaptive, personalized in a way that blurs the boundary between the two. AI content generation is already in production across the category. Full AI-tutor substitution for parts of the instructional function is a 2-5 year horizon, gated less by the technology than by procurement and trust cycles inside schools and districts.