Embedded AI Tutors Beat Legacy Platforms
Oboe
The core problem for incumbents is that AI turns their old advantage, large libraries of answers and lessons, into something users can get faster and often free elsewhere. Chegg built around search traffic, paid homework help, and a big archive of study content, but AI chatbots changed student behavior and Google sent fewer clicks. Duolingo and Khan Academy have adapted more cleanly because AI sits inside a habit based product, not on top of a traffic dependent content business.
-
Chegg shows how hard the transition is when the legacy funnel breaks first. By October 28, 2025, it had cut 388 roles, about 45% of staff, and brought back Dan Rosensweig as CEO after saying AI and lower Google traffic had materially reduced traffic and revenue.
-
Duolingo used AI as a premium layer on an already sticky app. Max added GPT-4 powered features like Roleplay and Explain My Answer, but by January 1, 2026 Duolingo had already moved Explain My Answer from paid add on to free, showing how quickly AI features get competed down inside incumbent products.
-
Khan Academy took a different route, using its content library and nonprofit distribution to price Khanmigo at $4 per month for learners and families, with free teacher tools. That makes AI tutoring feel like a low cost extension of existing coursework, not a separate premium subscription.
From here, education platforms that win will be the ones where AI is the product experience, not a bolt on upsell or a wrapper around old content. The market is moving toward low cost, always on tutors embedded inside daily learning loops, with pricing and workflows that make standalone answer libraries look obsolete.