Schools Want to Teach AI — They Just Don't Have the People
Most schools aren't avoiding AI education on purpose — they're missing trained people, not interest. Here's what the 2026 data says is actually blocking schools, and what we're exploring as a possible way to help.
Ask most school administrators whether AI literacy matters, and they'll agree immediately. Ask them whether their school is actually teaching it well, and the conversation changes. The gap between those two answers isn't about willingness. It's about people — specifically, not having enough trained ones.
We don't yet run a formal schools program. This article isn't a pitch for an existing product. It's an honest look at what the 2026 data says is actually blocking AI education in schools, and why we think a mentorship-based approach — the same model we use with individual families — is worth exploring as a partnership model for schools that want to move faster than they currently can on their own.
The barrier isn't interest. It's training.
The numbers from 2026 are consistent across multiple independent surveys, and they all point the same direction:
- Education Week's May 2026 special report found 42% of teachers have received no generative-AI training at all, while only 9% describe their training as ongoing rather than a single session. Most PD, where it exists, is a one-off, not a sustained program.
- A Gallup survey conducted with the Walton Family Foundation ("Teaching for Tomorrow," Feb-Mar 2026, over 2,000 teachers) found only 18% of teachers have received formal, written guidance on AI from their administration — 34% have no guidance of any kind across the ten most common AI-related tasks measured, and the rest rely on informal, verbal norms that vary teacher to teacher.
- The gap isn't evenly distributed. RAND's research, reported in spring 2025, projected that nearly all low-poverty districts would have trained teachers by the start of the 2025-26 school year — compared to only 6 in 10 high-poverty districts. That's a projection, not a new 2026 finding, but it's the clearest evidence that under-resourced schools are falling behind fastest — the opposite of what good policy would produce.
None of this is a story about schools not caring. It's a story about schools being asked to teach something new without being given the people, time, or training to do it well.
What the numbers actually describe
Put plainly, the ask these surveys describe is narrow and specific — not more enthusiasm, not another policy memo. It's tools, training, and ongoing support, delivered by people who already have the depth of understanding most classroom teachers haven't had time to build, because they're busy doing the actual job of teaching everything else. The Gallup/Walton survey found this guidance gap is linked to higher reported teacher burnout — a sign that the ambiguity itself, not AI as a technology, is part of what's wearing teachers down.
Outside partnership is already a recognized fix — at every scale
Bringing in outside expertise for a gap a school can't fill internally isn't a new or unusual model — it's how schools have handled specialized instruction for decades. At the national level, you can see the pattern in 2026's highest-profile AI-education responses: ISTE, ASCD, and Google's three-year initiative aiming to reach 6 million U.S. educators; OpenAI Academy and the Walton Family Foundation's in-person training reaching 1,600+ K-12 leaders; Florida's $5.7 million state-university partnership targeting high-need districts. These are large, well-funded efforts, and we're not claiming to operate at that scale or to have their institutional track record.
What we'd point to instead is the policy signal underneath all of them: the K-12 AI Literacy and Readiness Act of 2026 (introduced in both the House and Senate in 2026, not yet passed) would let districts use existing federal funding to bring in outside AI instruction and training if it becomes law — an early sign that "partner with an outside expert, even a smaller one" is moving toward becoming a recognized, fundable path for a district, not just something only a Google-scale partner can offer. The pattern that matters for a school evaluating options isn't limited to national programs: it's that depth from an outside specialist, at whatever scale fits a school's actual need, beats asking already-overloaded teachers to become AI experts overnight.
Why we think mentorship specifically is worth exploring here
Our existing, proven model is 1:1 AI mentorship for individual kids and teens — a real mentor working with a student on real, finished projects, not a generic course or an unsupervised chatbot. We think the same core idea — depth over breadth, real projects over passive content, a person who can answer the question that comes up mid-lesson — is worth exploring as something a school could plug into at a scale that fits them, rather than something that only works bought at national scale.
That could look like a few different things, and we want to be upfront that none of this is a packaged product yet: mentor-led workshops for students that take pressure off classroom teachers, structured project-based sessions that give kids the "building, not just chatting" experience we already use with individual families, or training support for teachers themselves, closer to the explain-the-why model our mentors already use one-on-one.
Any version of this would need to meet the standard schools rightly expect from anyone working with students: background-checked, vetted mentors, and clear handling of student data consistent with FERPA and COPPA requirements, not just our existing consumer privacy practices. We'd expect any real conversation with a school to start there, not end there.
What we're confident about is the problem itself: schools don't need more conviction that AI literacy matters. They need people who can actually deliver it, at a scale that fits their budget and size, and right now, most don't have enough of them.
If you're a school administrator or teacher reading this
We'd genuinely like to hear from you — not to sell something that doesn't exist yet, but to understand what a useful partnership would actually need to look like from your side. What's the biggest blocker at your school: time, training, policy clarity, or something else? What would actually help, versus what sounds good in a pitch but wouldn't survive contact with a real classroom?
If you want to talk through what this could look like for your school, email us directly at [email protected] — not the general parent booking line, an actual conversation with no pitch and no assumption that we have the answer yet. We're trying to build the right thing, and that starts with understanding the problem from people actually living it.
FAQs
Q: Does Build AI With Us currently have a formal schools program? A: Not yet. Our existing, proven model is 1:1 mentorship for individual families. This article reflects our thinking about whether and how that model could extend to schools, based on real 2026 research into what's actually blocking AI education in classrooms — it's not an announcement of an existing product.
Q: Is the problem really a lack of teacher interest in AI? A: No — the data doesn't support that. Multiple 2026 surveys (EdWeek, Gallup/Walton) show the overwhelming majority of the gap is training and guidance, not interest or willingness. Teachers consistently describe wanting more support, not less involvement.
Q: Why would a mentorship model work better than a standard training course? A: We don't claim it definitively would — that's part of what we'd want to learn from actual schools. But our experience with individual students suggests that depth (real projects, a person to ask follow-up questions) produces more durable understanding than one-off training sessions, which is exactly the structural weakness a 2026 review of 70 studies found in most current school AI-PD efforts.
Q: Are there real examples of outside organizations successfully partnering with schools on AI education? A: Yes. ISTE/ASCD's partnership with Google, OpenAI Academy's work with the Walton Family Foundation, and Florida's state-university partnership through FAU are all active, named 2026 examples of this exact model working at scale.
Q: What should a school do if they're interested but nothing formal exists yet? A: Email [email protected] and start the conversation directly. The earliest, most useful partnerships usually start from an honest conversation about a school's specific gap, not from a fixed package being sold to them.
Q: How would student safety and data privacy be handled in a school partnership? A: We'd expect to meet the standard schools rightly require for anyone working with students — background-checked, vetted mentors, and student data handled consistent with FERPA and COPPA, not just our existing consumer privacy practices. This is a starting requirement for any real conversation with a school, not an afterthought.
Ready to see if 1:1 AI mentorship is right for your child?
Book a Free Trial Class. Meet a mentor and get a personalized path.
