Khan Academy has been quietly building toward this moment for a couple of years now. Starting this fall, students and teachers using Khan Academy will encounter a new generation of Gemini-powered classroom tools — interactive visual aids and teacher-controlled practice materials that Google says are ready for the new school year. This isn’t a pilot or a waitlist. It’s shipping. And depending on how it lands in actual classrooms, it could be one of the more meaningful AI deployments in K-12 education we’ve seen yet.
How Google and Khan Academy Got Here
The Google-Khan Academy relationship isn’t new. Khan Academy has been piloting AI tutoring through its Khanmigo platform since early 2023, initially powered by GPT-4. The shift toward Google’s Gemini models has been gradual but deliberate — Khan Academy’s nonprofit structure means they’re particularly sensitive to cost-per-query economics, and Google has been aggressive about offering favorable terms to education partners.
Sal Khan has been unusually candid about what he wants AI to do in schools. His vision — a personal Socratic tutor for every student on the planet — sounds ambitious, but Khan Academy has the scale to attempt it. The platform serves over 150 million registered users globally. Even a modest AI feature rollout touches millions of students.
The timing here also matters. Schools are heading into a fall semester where AI is no longer a novelty conversation in the faculty lounge — it’s a procurement decision, a curriculum question, and increasingly a parent concern. Google and Khan Academy are betting that showing up with polished, teacher-friendly tools right at the start of the school year is smarter than launching mid-semester when everyone’s already overwhelmed.
OpenAI has been moving in the same direction — ChatGPT for Teachers recently expanded to 55 more U.S. school districts — so this announcement is also Google staking its claim in a competition that’s becoming very real, very fast.
What’s Actually Launching: The Features Breakdown
Google’s official blog post on the Khan Academy partnership highlights two main product launches for this back-to-school season. Let’s break them down properly.
Interactive Visual Aids Powered by Gemini
The first is a set of Gemini-generated interactive visual aids built directly into Khan Academy’s learning interface. These aren’t static diagrams pulled from a library. The idea is that Gemini generates contextual visual explanations — think dynamic graphs, animated concept walkthroughs, illustrated problem breakdowns — tailored to what a student is actually working on.
This is meaningful for subjects like geometry, physics, and biology where text explanations only go so far. A student stuck on understanding how a sine wave behaves doesn’t just need more words. They need to see it move. If Gemini can generate that on-demand rather than relying on pre-made assets, the content coverage expands dramatically.
The multimodal capabilities of Gemini — specifically its ability to reason across text and images simultaneously — are what make this technically feasible. Earlier AI tutoring systems struggled here because they were essentially language models bolted onto static content libraries. Gemini’s architecture handles visual reasoning more natively.
Teacher-Controlled Practice Materials
The second launch is arguably more immediately practical for schools: teacher-controlled AI-generated practice materials. Teachers can use Gemini through Khan Academy’s platform to generate customized problem sets, quizzes, and exercises aligned to what their class is actually covering — not just the standard curriculum sequence.
This addresses a real pain point. Teachers spend enormous amounts of time creating differentiated materials for students at different levels. A tool that lets a teacher say “give me 10 word problems on two-digit multiplication for students who are struggling with carrying” and actually get something usable — that’s genuinely useful, not just impressive in a demo.
Key details about both launches:
- Both tools are built on Google’s Gemini models, with Khan Academy integrating via Google’s API
- Teacher accounts retain control over what AI-generated content gets assigned to students
- The tools are launching for the 2026-2027 school year, available now to Khan Academy users
- Khan Academy’s existing Khanmigo AI tutor continues alongside these new features
- Availability is initially focused on U.S. classrooms, with international rollout details unconfirmed
- Khan Academy remains free for students; teacher features are part of Khan Academy for Teachers at no cost
What This Means for Khanmigo
It’s worth understanding how these new tools sit alongside Khanmigo. Khanmigo is Khan Academy’s conversational AI tutor — a chat-based interface that helps students work through problems using the Socratic method rather than just handing over answers. The new Gemini-powered features seem designed to complement Khanmigo rather than replace it: Khanmigo handles the tutoring dialogue, while the new visual aids and practice generators add richer content scaffolding around that conversation.
Whether that integration feels cohesive to students in practice, or whether it creates a fragmented experience, is something we won’t know until classrooms actually use it this fall.
Why This Matters — and Who It Actually Helps
There’s a real tendency in EdTech to announce things that sound useful and then quietly disappear into the void of school IT procurement nightmares. So let’s be honest about who this actually benefits and where the friction points are.
Teachers Get the Most Immediate Value
The practice material generator is the feature most likely to see adoption, because it solves a problem teachers feel every single week. Curriculum differentiation is time-consuming. If an AI can produce a decent first draft of a problem set in 30 seconds that a teacher then edits rather than creates from scratch, that’s real time saved. The teacher-control framing is also smart — it addresses the understandable anxiety about AI just dumping inappropriate or inaccurate content in front of kids without a human checkpoint.
Our earlier look at what a 1,000-student study found about ChatGPT and critical thinking should give pause here, though. There’s a non-trivial concern that AI-generated practice materials, if not thoughtfully designed, might optimize for completion rather than comprehension. Teachers using this tool will need to evaluate not just whether the problems are correct, but whether they’re pedagogically useful.
Students Get Richer Explanations — In Theory
The interactive visuals are exciting on paper. Dynamic, on-demand visual explanations could genuinely help students who struggle with abstract concepts. But this is also where AI tools most often overpromise. The quality of AI-generated visuals varies, and in an educational context, an incorrect or misleading diagram is worse than no diagram at all. Khan Academy’s team will need robust quality controls here.
Google’s Competitive Play Is Obvious
Let’s not pretend this is purely altruistic. Google is building a case for Gemini in education at a moment when Gemini in Google Workspace is already pushing into schools. Locking in the Khan Academy partnership — with its massive global reach and trusted brand — is a significant strategic move. If students and teachers build habits around Gemini-powered tools during their formative educational years, that has long-term implications for Google’s user base.
OpenAI knows this too. Microsoft’s partnership with education platforms, OpenAI’s direct school district deals — this is a land grab, and Khan Academy is one of the most valuable pieces of territory on the board.
What This Means for Different Users
Here’s how to think about the practical impact depending on where you sit:
- Teachers: The practice material generator is worth trying immediately. Approach it as a drafting tool, not a finished product. Review everything before assigning it.
- Students: The visual aids could genuinely help when you’re stuck on a concept — treat them as one explanation among several, not the definitive answer.
- School administrators: This costs nothing extra on top of existing Khan Academy access, which removes a major barrier. The data privacy question is still worth asking your IT team.
- Parents: Khan Academy’s teacher-control feature means a human is still in the loop on what AI content reaches your child. That’s the right design, but hold schools accountable for actually using those controls.
FAQ
What exactly are the new Gemini-powered tools Khan Academy is launching?
Khan Academy is launching two main features: interactive visual aids generated by Gemini that help explain concepts dynamically, and a tool that lets teachers generate customized AI-powered practice materials for their students. Both are built using Google’s Gemini models and are available for the 2026-2027 school year.
Is this free to use?
Khan Academy is free for students, and its teacher tools have traditionally been offered at no cost through Khan Academy for Teachers. The new Gemini-powered features appear to follow the same model, though it’s worth checking Khan Academy’s current offerings directly on their site for the most up-to-date details.
How does this compare to what OpenAI is doing in education?
OpenAI has been expanding ChatGPT access to school districts across the U.S. and building teacher-specific features, making this a direct competitive space. Google’s advantage here is the existing Khan Academy partnership and the platform’s trusted reputation in K-12 education globally — Khan Academy carries more institutional trust with schools than ChatGPT does right now.
What about student data privacy?
This is the question schools should be asking before deploying any AI tool. Khan Academy has a strong track record on student data privacy and complies with COPPA and FERPA. The Google integration adds a layer worth scrutinizing — schools should review Khan Academy’s updated data processing agreements to understand exactly how Gemini handles student interactions.
The fall semester will be the real test. Announcements are easy; classroom adoption is hard, messy, and full of surprises that no product demo anticipates. If these tools hold up under the chaos of actual school days — inconsistent internet, underprepared teachers, students who immediately try to break the system — then Google and Khan Academy will have built something worth talking about for years. I’d expect a clearer picture by October, once teachers have had a few weeks to either integrate these tools into their routines or quietly stop using them.
