AI Tutors Are Here, But Student Engagement Lags
AI tutoring adoption is not matching the hype. Khan Academy’s Khanmigo has logged more than 108 million interactions since its rollout in 2023, yet the organisation recently acknowledged that only about 15 percent of students with access to the tool use it regularly. That gap highlights a core issue for classroom AI tools: availability does not guarantee engagement. Early classroom deployments produced inconsistent results, prompting Khan Academy to rework Khanmigo and rethink how it is woven into everyday learning. Instead of assuming students will seek out a chatbot on their own, platforms are starting to ask tougher questions: Where does AI actually fit in a lesson? When do teachers trust it enough to recommend it? And how can it support, rather than distract from, existing teaching practices and mastery-based learning models?
Khan Academy’s Classroom Redesign: Embedding AI in Daily Workflows
To address low student AI engagement, Khan Academy has redesigned its classroom experience around teacher and student workflows. For teachers, a new dashboard centres on core tasks such as managing classes, assigning work, reviewing reports, and accessing AI tools without navigating multiple menus. The Khanmigo Assistant now sits at the top of this experience, allowing teachers to search for content and move around the platform using natural language. Khanmigo Teacher Tools support planning tasks, from lesson hooks to individualized education program support, keeping preparation inside the same environment where assignments are created. For students, a new Learner dashboard and structured Learner Queue translate assignments into clear daily or weekly Missions, broken into smaller steps. Motivation features like gems, weekly streaks, and Gem Challenges aim to make regular interaction—potentially including AI support—part of the normal rhythm of classwork rather than an optional extra.
Canva Learn Grid: Extending Classroom AI Tools Beyond School Walls
While Khan Academy is reshaping in-class workflows, Canva is expanding classroom AI tools outward with Learn Grid, a free platform for teachers, parents, tutors, home learners, and lifelong learners. Learn Grid combines more than 50,000 curriculum‑mapped resources with AI‑powered activity generation across over 30 activity types and supports more than 16 languages. Users can search by subject, grade, and learning outcome, then adapt or generate interactive games, worksheets, whiteboards, and graphic organisers directly in Canva. For verified teachers, Learn Grid integrates planning, assignment, live lesson delivery, and automatic student response data in one place, reducing the need to juggle multiple systems. For parents and tutors, it offers structured, aligned materials to support homework or targeted practice. By positioning AI inside the full cycle of discovering, creating, assigning, and reviewing learning activities, Canva aims to make AI a silent partner in lesson design rather than a standalone novelty.
From Optional Add‑Ons to Integrated Infrastructure
The strategic shift in both platforms points to a broader trend in educational platform redesign: AI is moving from add‑on feature to embedded infrastructure. Khan Academy’s revamped dashboards and learning queues are designed so that AI tutoring sits inside the same pathways teachers already use to assign content and track mastery. Canva’s Learn Grid similarly integrates AI into planning, activity creation, delivery, and feedback. This deeper integration reflects a recognition that student AI engagement rises when tools appear at the exact moment of need—during a lesson hook, a tricky exercise, or targeted practice—not on a separate tab. It also acknowledges that teachers, not just students, are primary users; if AI reduces their cognitive load, they are more likely to invite students to use it. The goal is less about showcasing AI and more about making everyday tasks smoother and more personalised.
The Human Barriers to AI Tutoring Adoption
Despite these redesigns, AI tutoring adoption still depends on human factors. Teacher trust is critical: educators need confidence that AI suggestions align with curriculum goals, support diverse learners, and do not undermine mastery‑based approaches. Khan Academy’s admission of inconsistent early results with Khanmigo underlines how easily skepticism can grow if AI output feels unreliable or poorly integrated with class objectives. Students, meanwhile, are unlikely to embrace AI simply because it is available. They respond to clear routines, visible benefits, engaging interfaces, and sometimes game‑like incentives such as gems or streaks. Canva’s framing of Learn Grid around real‑world pain points—like teachers spending Sunday evenings hunting for aligned worksheets—signals another key insight: solving immediate, practical problems is often more persuasive than abstract promises of personalisation. Ultimately, meaningful AI tutoring adoption will hinge on platforms aligning design, workflow, and trust with the realities of everyday teaching and learning.