Your child comes home with an assignment built using an AI tool, and you find yourself asking the question every parent now asks: did she learn this, or did the software do it for her? A University of Kansas program is helping teachers across the state answer a version of that question with a “stoplight method” that grades every assignment red, orange, yellow or green by how much AI a student is allowed to use. It is thoughtful work, it is spreading, and teachers are grateful for it. But the stoplight, like nearly every school AI policy being written right now, is built to answer a question about access. The question that decides whether a tool teaches your child anything is a different one, and almost nobody is asking it.
The spread of AI policies like the Kansas stoplight raises practical questions for parents deciding what to allow at home. Here are the ones coming up most.
Common questions
What is the “stoplight method” for AI in schools?
Should I let my child use AI for homework?
How do I tell if an educational app is actually teaching my child?
How do I figure out where my child actually needs help?
Kansas schools now sort assignments by how much AI is allowed. Sensible. But that was never the question that decides whether a tool teaches. The real one: is your child still doing the thinking?
What Kansas built
On September 3, KCUR reporter Rachel Schnelle profiled a program run by the University of Kansas Center for Reimagining Education, housed in the university’s Achievement & Assessment Institute. The center gives Kansas districts grant funding and professional development to help teachers work with AI. This year 11 schools received $7,000 stipends through its cohort model, and a free online module exists for districts that want the material without the stipend. Its centerpiece is the “stoplight method,” a four-level guide to how much AI a given assignment allows: red means no AI, orange means AI only at the end of the assignment, yellow makes AI a “study buddy,” and green treats it as a “co-pilot.”
In one classroom the piece describes, Heather Campbell, an algebra teacher at Bonner Springs High School, has students use a tool called MagicSchool to translate a math concept, slope-intercept form, into real-world jobs. She types in careers like electrician or anesthesiologist to show how the math turns up in each. What she wants is not copy-and-paste. “Anybody can do that,” she said. “But that thinking, and the going back and forth between AI, enhancing your voice, is the part that my goal is for this year.”
The framework is spreading. Districts from Bonner Springs to Circle Public Schools to Morris County are using it, and separate reporting found teachers at Jefferson West doing the same. Bart Swartz, who directs the center, describes the strategy in terms of adoption rather than pedagogy. “There are so many times that districts start on an initiative and they either start with the whole building or the whole district trying to take on something new and it often fizzles out because people don’t have buy-in,” he said.
Anybody can do that. But that thinking, and the going back and forth between AI, enhancing your voice, is the part that my goal is for this year.
Laura Lurns · Learning Success expert
The question the stoplight steps around
Look closely at what the stoplight sorts. Every level is about access and disclosure: how much AI a student is allowed on this assignment, and at what stage. That is a reasonable thing to sort out, and a teacher needs a shared vocabulary for it. But access is not what decides whether a tool teaches. The thing that decides that is whether the task still makes the child do the effortful work that builds a skill, or hands that work to the machine. The framing almost every school has adopted, the one that treats AI in class as a matter of permission and etiquette, steps right around it.
Here is where the science actually points, and it is worth being careful about what it does and does not say. Decades of learning research, from Roediger and Karpicke’s work on retrieval practice onward, point in one direction: memory that lasts is built by effortful recall, by practice spread out over time, and by feedback that names the next step rather than only marking an answer right. What makes a piece of software teach is whether it is built around those principles, not whether it is powered by AI or how engaging it looks. The honest limit is that none of this research tested generative-AI tools like MagicSchool in a real classroom. It hands a parent a test to apply, not a verdict on any one tool. This is not an argument against technology. It is an argument against technology that fakes the principle, that produces the look of learning without the work of it.
Apply the test to the stoplight and the colors come apart. A “green” assignment where a child asks a chatbot to draft the essay has outsourced the exact thinking that builds writing. A “red” one that has her retrieve and explain in her own words preserves it. The permission level tells you almost nothing about which is happening. This is the same trap ed-tech has fallen into for years, optimizing for how much a child uses a tool instead of whether they are getting more capable. A child’s interest and motivation are not the same as their learning, and a tool that holds attention beautifully might leave the underlying skill untouched. Our own position puts it plainly: a points system makes a child feel like they are learning, while real mastery makes them feel capable, and those are not the same thing. Campbell, notably, is already reaching for the right line. The back-and-forth, the thinking, the child enhancing her own voice is exactly the work worth protecting.
Key takeaways
- The stoplight: Kansas teachers now grade assignments red to green by how much AI a student is allowed.
- Wrong question: Access and disclosure rules skip the thing that decides whether a tool actually teaches.
- The test: Ask whether the task still makes your child do the effortful work, or hands it to the tool.
What to ask about the tools your child uses
The question worth carrying into a parent-teacher conference is not whether your child is allowed to use AI. It is the one our work keeps coming back to: is this tool building the skill, or replacing the effort that builds it? A practical version is even simpler. Ask what your child actually did that the tool did not. If she described the problem, wrestled with the back-and-forth, and revised the result in her own words, the effort that teaches is still hers. If the tool produced the answer and she moved it along, the work that builds the skill happened somewhere else, or nowhere.
Held to that test, AI in the classroom is neither a gift nor a threat on its own. Like the mandatory reading screeners several states have adopted, it helps or harms depending entirely on how it is used. A tool that quizzes a child, spaces out practice, or points out where their reasoning went wrong is on the right side of the line. One that drafts the essay, summarizes the chapter, or solves the problem on request is on the wrong side, however polished it looks. The color-coded permission level will not tell you which; the design of the task will. And it is worth saying plainly that even the tools built the right way are a supplement to good instruction and to your relationship with your child, not a substitute for either. No app replaces a teaching adult, and none replaces you.
The Kansas educators working out access rules are doing something useful, and the teachers reaching, like Campbell, for the thinking underneath are doing something better. The conversation the field has barely started is the harder one: not how much AI a child is permitted, but whether the task is designed so the child still does the learning. You do not have to wait for a committee to have that conversation. You are free to have it tonight, at your own kitchen table, over whatever your child brings home.
There are so many times that districts start on an initiative and they either start with the whole building or the whole district trying to take on something new and it often fizzles out because people don’t have buy-in.
Laura Lurns · Learning Success expert
You do not need a district committee to tell you whether your child is learning. You are the one who sees whether the work is building something or being handed over, and that is more real judgment than any permission chart holds. The villain here is not AI, and it is not the Kansas teachers thinking hard about it, who are asking a fair question and often reaching for a better one. It is the older, tidier idea that the AI question is about access, so a child who used the tool correctly is a child who learned. Learning Success was built to refuse that trade. Our All-Access membership opens an assessment that asks about the processing systems your child’s learning runs on, then a roadmap that names which skill to build first. It is a place to start helping your child today, not another tool to keep them busy.
See what All Access gives your child
See MTSS for schools
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.
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
