Education news

Washington Wants an AI Lab in Every School. What Actually Teaches?

A bipartisan bill would put an AI lab in every K-12 school, on a wave of AI-powered classroom products making the same promise. But decades of learning research point to an uncomfortable truth: what makes any technology teach is its design and the adults around it, not the label.

Laura Lurns Last updated September 7, 2026 · 9 min read

Washington Wants an AI Lab in Every School. What Actually Teaches?

If your child’s school sent home a proud note about a new AI lab, you probably felt two things at once: hope that technology will finally help, and a quiet worry that it is one more screen. On September 4, 2026, at a middle school in Ridgewood, New Jersey, Rep. Josh Gottheimer unveiled the AI LABS Act, a bipartisan proposal to fund dedicated AI labs in K-12 public schools, complete with curriculum and teacher training. It is a serious idea, and it arrives on a back-to-school wave of AI-powered learning products all making the same promise. Here is the quieter truth from decades of learning research: what makes any technology teach is how it is designed and taught, not the label on the box. The hardware was never the thing doing the teaching.

TL;DR

The AI LABS Act, unveiled September 4, 2026 at George Washington Middle School in Ridgewood, NJ, is co-led by Rep. Josh Gottheimer (D-NJ-5) and Rep. Jay Obernolte (R-CA-23); it would fund dedicated K-12 “AI labs” with curriculum and teacher training, and no bill number or funding figure appears in any public source yet.
It was announced alongside the bipartisan UNPLUGGED Act (Reps. Eugene Vindman D-VA-7 and Jen Kiggans R-VA-2), which supports phone-free schools and funds secure phone storage.
The OECD found that countries investing heavily in classroom technology saw no appreciable improvement in reading, mathematics, or science, a null result about benefit and not evidence that technology harmed learning.
Tutoring research (VanLehn, 2011) finds systems that respond to the steps of a student’s reasoning approach human-tutor effectiveness, while software that only checks the final answer barely helps; the design of the interaction, not the medium, is what matters.
The “45 percent” and “two-thirds cheating” figures cited in the announcement come from a Rutgers survey of adults and university students, not K-12 children, and no study has evaluated “AI labs” as a category.

Context the coverage leaves out

What the coverage gets wrong

Most coverage of AI-in-school announcements repeats a policymaker’s or a company’s own words back as if they were results, treating “AI-powered” and “personalized” as proof that learning is happening. Two things get lost. First, the evidence: buying classroom technology has not been shown to raise achievement (OECD), a null result about benefit that is routinely mis-told as either a triumph or a disaster, and is neither. Second, the numbers. The widely repeated statistics that “45 percent” of students use AI frequently and “two-thirds” admit using it to cheat trace to a Rutgers survey of adults and university students, not K-12 children, and one district’s parent survey is not a national finding. The question that actually decides whether an AI lab teaches, whether the tools work through a child’s reasoning and whether teachers are trained to use them, is the one the coverage skips.

An AI lab in every school sounds like progress, but the research is clear that the label matters far less than the design. Here are the questions parents are asking about what actually makes classroom technology worth a child’s time.

Common questions

Will an AI lab in my child’s school actually help them learn?

It depends entirely on how it is built and used. Buying classroom technology, on its own, has not been shown to raise achievement (OECD), though that is a null result about benefit and not proof of harm. What predicts whether a tool teaches is its design: whether it works through the steps of your child’s reasoning, asks them to recall and practice over time, and whether teachers are trained to build on it. The same “AI lab” label sits on top of either.

Is AI bad for kids’ learning, or good?

Neither, by itself. The technology is not the mechanism. Well-designed tools, step-level tutoring, retrieval-based practice, and feedback that tells a child their next move, genuinely help as a supplement to a teaching adult. Points, badges, and streaks are engagement features, and engagement is not the same as learning. The label on the box tells you what powers a tool, not whether it teaches.

How do I tell if a learning app is teaching my child or only keeping them busy?

Watch for ten minutes. When your child gets something wrong, does the tool show them the step they missed and ask them to try it again, or does it flash a red mark, hand out points, and move on? The first is teaching. The second is engagement. A tool could run on the most advanced AI available and still only grade answers.

How do I figure out what my child actually needs before trusting any tool or lab?

Start by getting clear on where your child is, in language that builds them up rather than boxing them in. A screener or the Learning Success assessment asks what you are seeing at home across the systems learning runs on, and the roadmap that comes back names what to build first. A screener is a starting point, not a diagnosis. If your child might need formal accommodations such as an IEP or 504 plan, or you suspect a vision, hearing, or medical cause, pursue a professional evaluation too, since that is the only route to those supports.

Share the idea

Washington wants an AI lab in every school. But research says what makes technology teach is how it is designed and taught, not the label on the box. The hardware was never the teacher.

Post on X Follow

What happened

On September 4, 2026, at George Washington Middle School in Ridgewood, New Jersey, Rep. Josh Gottheimer (D-NJ-5) unveiled the AI LABS Act, a bipartisan proposal he is co-leading with Rep. Jay Obernolte (R-CA-23). The bill would direct the U.S. Department of Education to fund dedicated “AI labs” in K-12 public schools, described as the modern version of the computer and typing labs of earlier decades, with curricula that teach how AI works and how to use it responsibly, plus training for teachers. Gottheimer pitched it as a third way between banning AI outright and, in his framing, drowning in it.

“AI can be a powerful tool, but it can never replace the fundamentals,” Gottheimer said. He announced it alongside a separate bipartisan measure, the UNPLUGGED Act, sponsored by Reps. Eugene Vindman (D-VA-7) and Jen Kiggans (R-VA-2), which would support phone-free school policies and fund secure phone storage, with exceptions for medical need, disability, and English-language learners.

Worth noting what the announcement does not yet include: as of this writing there is no bill number, no funding figure, and no evaluation requirement in any public source. This is a proposal that has been unveiled, not enacted or even formally filed. And unlike a pure hardware grant, the AI LABS Act names the two ingredients the evidence cares about most, curriculum and trained teachers, which makes the real question not what it buys but how any given school builds what it buys.

AI can be a powerful tool, but it can never replace the fundamentals.
Laura Lurns · Learning Success expert

The frame the science supports

The assumption riding underneath most AI-in-school coverage is that the technology is the teaching, so funding the lab and adding the word “AI” is itself the win. Decades of research point the other way. When the OECD studied countries that had invested heavily in classroom technology, it found no appreciable improvement in student achievement in reading, mathematics, or science. Read that carefully, because it is routinely mis-told in both directions: it is a null result about benefit, not evidence of harm. Buying classroom technology has not been shown to raise achievement. That is a statement about what money spent on devices, by itself, buys, and the AI LABS Act is smarter than that baseline precisely because it also funds curriculum and teacher training.

So what does make a screen teach? The design of the interaction, not the chip behind it. In Kurt VanLehn’s review of controlled tutoring studies, systems that worked through the individual steps of a student’s reasoning approached the effectiveness of a human tutor, while software that only checked the final answer barely moved the needle. Two of the most replicated findings in how people learn fill in the rest: pulling information out of memory beats rereading it (Roediger and Karpicke, 2006), and spacing that practice over time beats cramming it (Cepeda and colleagues, 2006). Feedback helps when it tells a child their next move, and backfires often enough when it only keeps score (Kluger and DeNisi, 1996). None of this is anti-technology. Lexia Core5, a screen-based reading program, is rated by the What Works Clearinghouse and holds up in independent evaluation, which is exactly the standard every “AI-powered” product should be held to.

Here is the honest catch, and it is the reason to slow down rather than cheer or jeer. No study anywhere has evaluated “AI labs” as a category, because the label describes a room, not a method. A lab stocked with drill-and-reward software and no follow-through for teachers fails in exactly the way the OECD case describes. A lab built on step-level, well-designed tools with trained educators is the VanLehn case. Same label, opposite outcomes, and the bill’s text decides none of it. The value lives entirely in implementation choices no headline mentions.

Key takeaways

The label is not the lesson: a lab teaches by how its tools are designed and taught, not by the word AI.
Spending is not the same as results: buying classroom technology has not been shown to raise achievement.
Implementation decides everything: the same AI lab helps or fails on the tools and whether teachers are trained.

What it means for your child

You do not need to grade an algorithm to make a good call. You need to know what teaching looks like and trust yourself to spot it. If your school opens an AI lab, ask the questions the research would ask. When a child gets something wrong, does the tool show them the step they missed and ask them to try it again, or does it flash a red mark, hand out points, and move on? Are teachers trained to build on what the tool surfaces, or is the lab a room children visit alone? Is the software asking them to recall and practice over time, or mostly to stay entertained? A points system makes a child feel like they are learning. Real mastery makes them feel capable, and that is the kind of confidence that grows from a child’s own evidence of getting better.

This is the California-screening lesson in a new outfit. A policy that sounds unambiguously good is only as good as how it is used, and a parent deserves to understand the decision rather than be handed a verdict. An AI lab could be the best thing to happen to your child’s school or an expensive room full of glorified flash cards, and the difference is not the funding. It is the design and the adults.

Which points at the variable no grant buys. The most powerful force in your child’s learning is a person who knows what to look for and stays in the room while they work. Well-designed tools earn their place as a support for that person, never a substitute for them. That is where the real work happens, with or without a lab down the hall.

You do not need to understand a neural network to choose well for your child, and you do not need to wait for Washington to decide what good learning looks like. Children learn from practice built around how they actually think, and from an adult who stays in the room while it is hard. The villain here is not AI, and it is not the lawmakers behind a genuinely bipartisan bill. It is the older and stubborner belief that funding the hardware and stamping it “AI” is the same as improving learning. That belief is what Learning Success was built to refuse. Our All-Access membership opens an assessment that asks about the systems your child’s learning runs on, and a roadmap that names which skill to build first, with you coaching the practice that makes it stick.

See what All Access gives your child

Is your child struggling in school?

Get your free personalized learning roadmap

You describe what you see at home. We turn it into a plan you start this week.

Answer 5 short questionnaires about what you already notice, 30–45 minutes at your own kitchen table
Your child sits no test and gets no score: nothing to schedule, nothing for them to dread
You do the answering, the AI does the writing, and a person reviews it before it reaches you
Access all 40+ courses instantly: reading, math, focus, processing and more, with new ones added regularly

Why we use AI, plainly: it writes from a knowledge base our team maintains and audits. We work through it line by line and pull anything the evidence stops supporting. The roadmap you get on Tuesday reflects what we corrected on Monday, and a human still reads it before you do.

Your school district must evaluate your child free of charge if you ask in writing, whatever your income and whatever the outcome (US, 34 CFR 300.111 and 300.301(b)). That route takes time and answers a different question than you do. This one starts today, from what you already know.

Get my free assessment & 7-day trial Cancel anytime during your trial.

Your answers stay yours. We do not sell your personal information, and we do not hand identifiable assessment data to outside AI companies to train their models.

A screener is a starting point, not a diagnosis. If your child might need formal accommodations (an IEP or 504 plan), or you suspect a vision, hearing or medical cause, pursue a professional evaluation too. That is the only route to those supports.

References

Rep. Josh Gottheimer (primary) — Gottheimer Unveils Bipartisan Back-to-School Agenda for AI in K-12 Education
WRNJ Radio — Gottheimer unveils bipartisan AI education agenda for K-12 schools
The Ridgewood Blog — Rep. Gottheimer Unveils Bipartisan Plan for K-12 AI Labs and Phone-Free Classrooms
Rutgers University — Report Finds Broad Adoption of AI in New Jersey and Strong Support for Regulation
Montclair Local — Montclair Board of Education Weighs New Tech, AI Policies as Community Survey Highlights Concerns
Google — AI tools for students and educators: Back to School 2026
OECD (2015) — Students, Computers and Learning: Making the Connection
Kurt VanLehn (2011), Educational Psychologist 46(4):197–221 — The Relative Effectiveness of Human Tutoring, Intelligent Tutoring Systems, and Other Tutoring Systems
Kulik & Fletcher (2016), Review of Educational Research 86(1):42–78 — Effectiveness of Intelligent Tutoring Systems: A Meta-Analytic Review
Roediger & Karpicke (2006), Psychological Science 17(3):249–255 — Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention
Cepeda, Pashler, Vul, Wixted & Rohrer (2006), Psychological Bulletin 132(3):354–380 — Distributed Practice in Verbal Recall Tasks: A Review and Quantitative Synthesis

Laura Lurns · Learning Success expert Writes about the learning brain for parents who want plain answers. Every article is grounded in current neuroscience and classroom practice.