Beyond Anthropic and OpenAI: Google, Khan Academy, and the Rest of the AI-in-Education Landscape

Anthropic and OpenAI get most of the headlines, but they're not the only serious efforts in AI-for-education right now — and honestly, the more interesting story is how much these organisations are actively partnering with each other and with teacher unions, rather than just competing. Here's a fast tour of the rest of the landscape.

Google + Khan Academy: Khanmigo, now built on Gemini

Google and Khan Academy announced a partnership to bring Gemini's models into Khan Academy's Khanmigo tools. The first piece is a Writing Coach for grades 7–12 (in beta for grades 5–6) that guides students through outlining, drafting, and revising their own essays rather than generating one for them — the same guide-don't-give-away principle you'll see across most of these tools. A Reading Coach for grades 5–12 is expected to follow, letting teachers assign interactive readings with comprehension check-ins and class-level insight into where students are stuck.

On the teacher side, Khanmigo already offers standards-aligned lesson planning, quick summaries of recent student work, and auto-generated rubrics and exit tickets — and Khan Academy has stated it intends to keep this free for all US teachers.

ISTE+ASCD: the standards body weighing in

ISTE (the International Society for Technology in Education, now merged with ASCD) isn't building an AI product — it's the group most US schools already look to for technology standards, and it's been an explicit partner in both the OpenAI and industry-wide AI-fluency conversations. Its CEO's public position is that thoughtful training is what determines whether these tools help or just add noise — a useful reminder that the tool matters less than how a school rolls it out.

The American Federation of Teachers: one union, two AI partnerships

Notably, the AFT — one of the largest US teacher unions — is publicly working with both Anthropic (on a shared K-12 safety and privacy "Gold Standard") and OpenAI/Microsoft (on a $23 million, 400,000-educator training initiative). That's worth sitting with: the union isn't picking a side, it's trying to make sure teacher-protective standards exist regardless of which company's tool ends up in a given classroom.

What this means if you're choosing between tools

For a parent or teacher trying to get oriented, the practical takeaway isn't "which company is best" — it's that every major serious offering right now shares the same three commitments: don't train on student data, align to FERPA, and put the teacher (not the student unsupervised, and not the AI alone) in the driver's seat. If a tool your school is considering doesn't clearly state those three things, that's worth asking about before adoption, regardless of the vendor.