Wealthy families are paying up to $75,000 a year to enroll children in AI-powered private schools like Alpha School, which plans to expand to 50 campuses this year with more than 1,200 students already enrolled. The trend arrives as public school systems grapple with whether to restrict or embrace generative AI, and as foundational literacy scores for U.S. teenagers hit their lowest levels since international testing began in 2000.
Alpha School students spend two hours each morning on core academic subjects using an AI tutor. Afternoons shift to workshops focused on public speaking, relationship building, and other non-academic skills. Human staff called "Guides" remain present to motivate students and build relationships, but they do not design lessons or grade assignments.
"I definitely think using AI well is a skill in the future, if not the skill, so it's kind of a disservice not to have that in school," Sarah Cone, a venture capitalist whose 8-year-old daughter attends Alpha's New York location, told Fortune.
The split-screen reality of AI in schools
The same week Alpha expands, New York City announced a one-year ban on student-facing generative AI for children in 2-K through eighth grade. The policy whiplash reflects a broader uncertainty. Researchers have warned that indiscriminate AI use can erode the cognitive effort required for learning, yet studies also show that intentionally designed tools-such as chatbots that act as math tutors rather than answer dispensers-can improve outcomes.
That kind of thoughtful integration is harder to pull off in under-resourced public schools. Huriya Jabbar, a professor of education policy at the University of Southern California, pointed to high teacher turnover as a barrier. A school might train someone on how to engage students with AI, only to see them leave within a year. Schools serving low-income students also face greater pressure to raise test scores, leaving less bandwidth for experimenting with new instructional approaches.
"AI-just like any other tool, technology, curriculum-is going to play out in an already unequal landscape," Jabbar said.
What the data shows about screens and skills
The push to put more technology in classrooms is not new, but the large-scale rollout of computers in schools since 2014 is now being blamed for weakening students' ability to think. Neuroscientist Jared Cooney Horvath testified to Congress that screen-based learning distracts from deep cognitive work and depresses test scores. "Our kids are less cognitively capable than we were at their age," he said during the hearing.
AI is accelerating these dynamics. A 2025 College Board survey found that 84% of high school students use AI for homework, while one in five high schools permit AI use without any formal policy. The consequences are measurable. A study of 26,811 Chinese students in grades seven through 12 found that AI use boosted homework scores by 18%-then caused exam scores to drop by 20%.
Foundational literacy is deteriorating in parallel. Reading and math scores for 15-year-olds across 38 free-market democracies on the Programme for International Student Assessment hit their lowest average on record. U.S. students' average reading scores fell 14 points to 490, the lowest since the country joined the exam. The problem follows students into higher education, where professors report undergraduates arriving unable to process long passages and relying on AI-generated summaries. Gen Z Americans between 18 and 25 read an average of 5.8 books in 2025, compared to 12.1 for those 65 and older.
What literacy demands in an AI era
Kirsten Peterson, a senior project manager at the nonprofit Education Development Center, said the goal of developing strong readers and writers does not disappear just because AI can interpret texts and generate prose. "We want readers who can interpret, question, make connections, recognize perspective, evaluate evidence, sit with ambiguity, and form their own understanding," she said. "We have to be really careful not to use AI to remove the very thinking students need to practice in order to become literate."
For educators navigating this shift, structured professional development is becoming essential. The AI for Teachers Courses offer practical guidance on integrating the technology without sacrificing the cognitive work that builds literacy. Broader resources on AI for Education also address how schools can adopt tools intentionally rather than reactively.
Why this matters for educators
The Alpha School model shows what happens when AI adoption is paired with significant resources and low student-to-guide ratios. But the data on homework scores versus exam performance makes clear that access to AI without pedagogical structure backfires. For teachers and administrators, the immediate task is not deciding whether AI belongs in classrooms-it is already there, with 84% of high schoolers using it-but building the frameworks that channel it toward learning rather than shortcutting. The students who lose most are those whose schools lack the stability and training to implement AI with intention.
Your membership also unlocks:
