There is a powerful myth that if children don’t use AI in schools, our children will be “left behind.” That “equity” for students of color means every child should have unlimited access to a chatbot, a platform, and a virtual tutor.
Equity means ensuring every child has access to what research tells us supports learning, not flooding classrooms with surveillance technology that produces cognitive and emotional debt that will disproportionately harm students of color. As a neuroscientist, educator and parent of Afro-Latine boys, I have been carefully following these studies and observing my students’ performance. At this moment, a moratorium is the only answer to address the impact AI could have on our children.
All the existing independent research on AI and the brain suggests that AI use inhibits our ability to learn . As I tell my medical school students, learning requires the brain to retrieve information, evaluate evidence, tolerate uncertainty, revise thinking, and construct new understanding. When those cognitive processes are routinely outsourced to AI, we weaken the very capacities schools are supposed to develop.
AI in schools risks an entire generation of students becoming increasingly dependent on AI to read, write, solve problems, and make decisions instead of trusting and strengthening their own reasoning, critical thinking and creativity. Students of color deserve to learn to trust themselves, to think expansively and critically, to delight in their creativity and to love reading, learning and writing.
AI also reinforces structural racism and bias. I’ve heard countless stories about programs currently used in schools that fail to recognize accents or racially profile students. Tiera Tanksley argues that educational technologies reproduce structural racism because they inherit the inequities embedded within the institutions that create them. Similarly, Roderick Crooks has demonstrated that Ed Tech’s algorithmic systems are not race-neutral. They reflect existing power structures and amplify racial inequities under the appearance of objectivity.
Artificial intelligence cannot replace what decades of evidence have shown: Our children need small classes, ample resources, and caring teachers that affirm them and set high expectations. As a neuroscientist I know that children learn through relationships, experiences, inquiry, mistakes and productive struggle: the opposite of outsourcing thinking to AI. We should be expanding the already limited options of public progressive schools for children like mine, that describe their schools as places that they feel seen, heard and supported.
In communities like mine in East Harlem, equity should mean that our children are no longer over-policed, over-surveilled and over-tested. We want our Black children to have the appropriate learning environments with the support they need to thrive.
The people clamoring for AI in schools want an educational model that puts profit over humanity, compliance over curiosity, and surveillance over democracy. If we are not careful, AI will not dismantle the school-to-prison pipeline — it will supercharge it.
We need to actively fight against anything that harms our children. A two-year pause on AI in schools would give us time to make research-informed decisions, to understand AI’s impact on the brain, to thoroughly evaluate AI programs for racial bias, and to ensure that we’re spending public dollars only on evidence-based practices that help students be their best, most fulfilled selves.
Salas-Ramirez, a neuroscientist and associate professor at Rutgers University – Robert Wood Johnson Medical School, serves as the president of the Community Education Council for District 4, is a former member of the Panel for Education Policy and is a current public school parent of two boys.