In May, the Educational Opportunity Project at Stanford published test score data showing a decline in American students’ standardized test scores. The data show what teachers have long known: students spend too much time on screens and not enough time reading, writing, and thinking. This generation was handed a Chromebook in kindergarten and a cell phone before the end of elementary school, and now we are giving them unfettered access to generative AI.
Over the last 20 years, I have watched as my students’ willingness to focus on reading and writing dwindled. Very few students arrive in my 9th grade classroom as readers; most tell me they find reading boring. Even fewer are interested in writing. They do, however, arrive with personal laptops and smartphones. Fueled by constant dopamine hits thanks to unmitigated access to screens, they fail to see the value in difficult tasks like reading and writing.
According to the Pew Research Center, half of American teens spend most of their day online. I collect cell phones at the beginning of class, and I watch screens light up almost constantly with notifications from social media apps like Snapchat and Discord, text messages from family and friends, pop-ups from games like Cookie Clicker and Fortnite, and even alerts from gambling apps. Students receive dozens to a hundreds of notifications during a single class.
How can any curriculum compete?
We are starting to see tech remorse from school districts and parents, with schools cutting back on screen time and some parents restricting smartphone access at home. But at the same time, the use of generative AI in schools is increasing. Teachers are told they must use AI in their lessons, and students use AI to complete homework assignments, write essays, or create art (whether they are allowed to or not). Professional organizations like NCTE (National Council of Teachers of English) and NCSS (National Council of the Social Studies) have partnered with corporations like Google and IBM to promote AI in classrooms. Despite recent studies showing that generative AI use is detrimental to young minds, those outside the classroom insist it’s the only way forward.
It isn’t true.
K-12 students do not need access to generative AI in school.
Shae O. Omonijo, PhD candidate and member of the Harvard Innovation Lab, is just one expert who has called for a moratorium on K-12 access to AI. She points out that the people trying to push AI into schools didn’t learn to code or build apps in school- they learned it at home, through Youtube, hack events, and other informal education spaces. Why should we treat AI differently? She says, “…by the time they’ll “need it” it’ll be so far advanced that whatever you spent classroom time teaching them will be obsolete.”
It’s time to listen to educators, not edtech companies. It’s time to listen to the people who spend every day in school with students, not Silicon Valley CEOs with a financial interest in getting children addicted to screens and apps.
While many people view school as preparation for college and careers, school is not a workplace, and students are not employees. There are use cases for AI in education- data analysis, translation for parent communication, and differentiating reading levels- but in most cases, it should be treated like the single computer I had in my classroom in 2005. It should help adults; it’s not a thought partner for K-12 students.
Why are we so quick to refer to genAI as a thought partner rather than what it actually is- a probability machine? If we changed the terminology, our perspective on using it with students would be different. You can collaborate with a thought partner because it’s trustworthy (like a human); you use a predictive language machine to see what the most likely arrangement of letters and words might be.
Calling a predictive letter machine a “thought partner” or equivalent anthropomorphized term grants it a level of human trust. Adults struggle to understand that AI tools aren’t the same as interacting with a human, so how can we expect students not to fall into the same trap? If we routinely anthropomorphize tools like Gemini or Claude or GPT, students are less likely to fact-check the output or check it for bias.
When we treat generative AI as a tool capable of thinking and interacting like a human, students see no reason not to outsource their thinking to the tool. Instead of productively struggling with an academic text, they upload it to a genAI tool and ask for a summary. Instead of using their brains to come up with claims or thesis statements, they are “talking” to genAI and submitting its responses to teachers. They take photos of worksheets and assessments, upload them to genAI, and copy down the answers without thinking.
And yet we continue to see districts race to be the first to adopt AI into their curriculum.
Why are we gambling with a generation’s capacity to think? So we can be first in the AI arms race? So students can be prepared to prompt an AI tool that likely won’t even exist when they graduate high school or college?
If we want American students to be successful adults, they need the freedom to read, write, and think in K-12 classrooms. Teachers need to be able to challenge students to read and write regularly, to form opinions and communicate with others, and to practice executive function skills such as focus and stamina. GenAI is like a bulldozer that crashes through and dumps the answer in students’ laps.
Study after study has shown that offloading thinking to AI is overall harmful to children and teens. We know that constant screen time and access to social media are detrimental to young minds, but it is clear that toothpaste can’t be put back in the tube. With genAI, we still have time.
“But we must make them workforce-ready!”
Do we? In that case, shouldn’t we be replace English with “Slack and Salesforce for Sophs”? Maybe replace history with “Fun with Teams for Froshies”? GenAI is a tool, just like many workplace tools. Until now, we’ve trusted that students can adapt to employer training because of the skills they learn in K-12 education, not the specialized software covered.
The other argument tech evangelists make focuses on teaching students how to write prompts. Even the draft Working ELA AI Framework from NCTE and Google encourages teachers to incorporate prompting into the curriculum. Yet prompting has changed drastically in the last 18-24 months. It’s like teaching students how to create a Geocities website and telling them HTML will be vital to all careers in the future.
Our job as K-12 educators is to help students hone their abilities to read, write, and think critically. We have to give them opportunities for friction, for productive struggle, as they move through our classes. They need to feel ownership of their success and failures.
The AI literacy schools should be focusing on is how often genAI is wrong, incomplete, and biased. It’s not a magic answer machine that can be implicitly trusted. It’s not even an especially talented creator right now.
Generative AI should be banned from K-12 classrooms. If newsrooms are banning AI-created content, why are K-12 classrooms embracing it? Why are professional education organizations pushing teachers to use AI with students? We need to hit pause.
We need time and space to address student privacy and the right to their words.Your classmate or teacher won’t train on your writing or use it without your permission; with genAI, there’s no guarantee student writing will be protected, yet districts are requiring students to upload their writing to these tools for feedback! We’ve seen this script play out before with edtech tools like iReady, which is currently facing lawsuits over student privacy data and collection.Will AI tools face similar lawsuits?
We need time and space to address the effects of AI on young minds. According to Anne Maheux, assistant professor of psychology and neuroscience in the UNC College of Arts and Sciences and Winston Family Distinguished Fellow at the Winston Center on Technology and Brain Development, we are adopting AI at a speed that far outpaces the research being done on its effects. She points out teens are adopting its use far faster than they adopted the internet and social media, yet we haven’t learned from those experiences.
Educators and parents must demand a moratorium on genAI tools in K-12 schools. Just like we don’t allow minors to drink, drive a car, or join the army because their frontal lobes are not fully developed, we should protect them from genAI until we better understand its effects. It’s time to hit pause on AI in schools; the money spent on edtech tools can be better spent on purchasing books, field trip experiences, and other tangible items shown to help kids learn. Let them read novels, write stories, and draw pictures. Even if it’s hard, or they think they aren’t good at it, it’s real and it’s theirs. AI will be there later, when they have a strong foundation in the skills needed to be global citizens.
Filed under: miscellaneous | 2 Comments »
The Reading Zone 
