A cluster of AI-first private schools is charging wealthy American families tens of thousands of dollars per year to hand their children's early education to chatbots. Alpha School's kindergarten program runs $75,000 a year, and Silicon Valley venture capitalists are among the earliest customers. Neither Alpha School nor rival Forge Prep publishes performance metrics for its AI-guided model.
The pitch is that AI tutors, combined with what the schools call interactive project-based workshops, can replace traditional classroom instruction. Alpha School's footprint extends beyond kindergarten — some campuses run through high school at similar tuition tiers. Forge Prep operates on a comparable premise, marketing AI-assisted curricula to families willing to pay private-school prices for a product with no track record.
Shaun Johnson, a San Francisco-based venture capitalist, told the Wall Street Journal he plans to enroll his son in the $75,000-a-year Alpha Kindergarten. Johnson's rationale reflects a common framing among the schools' backers: that traditional K-12 is broken and that AI tutors are worth trying because the status quo has failed.
“We recognize that education is likely broken the way it is and there's going to be entrepreneurs that try to fix it… You want someone to be able to think on their feet and navigate the world, not necessarily a recitation of facts in a particular discipline.”— Shaun Johnson, San Francisco venture capitalist
Key facts
- 01Alpha Kindergarten charges $75,000 per year for an AI-tutor-led classroom, with tuition scaling through high school at some Alpha campuses.
- 02Forge Prep and Alpha School are marketing AI tutors and project-based workshops to families willing to pay tens of thousands per year.
- 03Alpha School co-founder MacKenzie Price says she plans to keep 'hot-button social issues' out of the classroom.
- 04Neither Forge Prep nor Alpha School publishes performance metrics, leaving no evidence the AI-guided model improves educational outcomes.
- 05Silicon Valley families are among the earliest adopters, including VC Shaun Johnson, who plans to enroll his son.
The economics of the segment are unusual. A $75,000 kindergarten tuition sits above most elite East Coast private schools, and the schools are selling access to a product category — AI-tutor-led classrooms — that has existed at scale for less than three years. Families paying that premium are effectively subsidizing product development on their own children, without the outcome data that a traditional private school would present to justify the price.
Alpha School co-founder MacKenzie Price has said the company plans to keep 'hot-button social issues' out of the classroom. What that excludes in practice is not spelled out, and the schools have not published curriculum standards or a scope-of-instruction document that would let outside educators evaluate the approach against state or accreditation benchmarks.
The transparency gap is the segment's central problem. Traditional private schools compete on measurable outputs — standardized test scores, college placement rates, AP results. Neither Forge Prep nor Alpha School releases equivalent data, which means the schools are asking families to underwrite a bet on AI-driven pedagogy with no visible way to check whether the bet is paying off. Independent researchers have not published peer-reviewed evaluations of either program.
There is a real research question underneath the marketing. Adaptive AI tutoring — Khan Academy's Khanmigo pilots, Carnegie Learning's math products, university-scale trials of tutor bots — has produced measurable learning gains in narrow, well-defined domains like algebra practice and language drilling. Whether those gains generalize to a full K-12 replacement, delivered as the primary instructional layer rather than as supplemental practice, is unresolved. Alpha and Forge are running that experiment live, at $75,000 a seat.
The customer base skews toward tech-industry families who are professionally comfortable with beta software and who tend to view educational disruption as directionally correct. That selection effect matters: the students most likely to succeed in an untested learning environment are also the students most likely to succeed in almost any environment, which will make it harder to isolate whether the AI tutor is doing the work or the household is.
Sycophancy is a specific concern with consumer LLMs used as tutors. Frontier models routinely validate incorrect user assertions, hedge on factual questions, and adjust their answers based on user pushback rather than evidence. A tutor bot that agrees with a seven-year-old's wrong answer is a worse teacher than a human who corrects the child, and the schools have not publicly addressed how their systems handle that failure mode.
The near-term business risk for Alpha and Forge is that the first cohort of graduates will eventually generate outcome data, whether the schools want to publish it or not. Once college admissions results, standardized test percentiles, and post-secondary trajectories become visible for children who spent K-12 with AI tutors as their primary instructors, the segment will either validate its premium pricing or collapse under it. Until then, the schools are selling optimism, and a very specific slice of the wealthy is buying.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
