Have you considered that the greatest risk of artificial intelligence in modern education might not be plagiarism, but the quiet, incremental erosion of students’ ability to think for themselves?
“How does Kate Chopin use irony to challenge the reader’s understanding of freedom?” This is the question asked by an educator standing in front of an 11th-grade English class. Normally, this serves as the cue for students to flip through their texts, underline a few lines, exchange initial ideas with classmates, and slowly begin constructing an analytical answer. Instead, dozens of pairs of blank eyes stare back. Shoulders shrug in unison.
“I don’t know,” one student offers.
Then another follows with a response that educators are hearing with increasing frequency in classrooms across the country: “I use ChatGPT. How am I supposed to know that?”
The room becomes quiet again—not because the students are deep in thought, but because many are not even trying. And therein lies the core dilemma.
Educators who have spent more than a decade teaching English and special education in major urban school districts like New York City have worked with countless students whose disabilities affect their reading and writing abilities, executive functioning, attention spans, and language processing. Over years of practice, teachers design specialized lessons that help students access learning without stripping away the vital cognitive work required to truly engage with the material.
The fundamental role of an educator has never been to lower expectations, but rather to remove unnecessary barriers while preserving academic rigor. Students have always struggled with complex concepts; that friction is an inherent part of the learning process. Lately, however, teachers are observing a distinct shift. Students are not merely struggling with challenging literature; they are increasingly struggling with the very act of thinking itself.
This crisis is not happening in a vacuum, nor did artificial intelligence single-handedly cause it. National reading data shows persistent, long-term declines in literacy achievement among adolescents. Writing research consistently demonstrates that high school students struggle most with analytical and evidence-based writing—essential tasks that cultivate deep critical thinking. While AI did not create these foundational vulnerabilities, it arrived precisely at a moment when those fundamental skills were already fragile. If educators are not intentional about encouraging students to wrestle with ideas, AI risks becoming a permanent substitute for the cognitive heavy lifting students most need to practice.
When AI Becomes the Brain Instead of the Thought Partner
Consider a common modern scenario: A student is absent from school, and an assignment is sent home. The next morning, the completed work arrives back at school, bearing little resemblance to the work that same student typically produces inside the classroom. The vocabulary is unusually sophisticated. The literary analysis flows effortlessly, complete with complex symbolism, even though the student struggled just the day before to identify a basic protagonist’s motivation.
Then, an in-class assessment is administered.
Working with a traditional pencil and paper, many of those exact same students visibly struggle to organize their ideas, support basic claims with textual evidence, or sustain a multi-paragraph response. This stark discrepancy goes far beyond simple cheating. It raises a much more fundamental question: Whose thinking is actually being measured and developed?
Researchers describe this phenomenon as cognitive offloading—the human tendency to rely on external tools to reduce mental effort. To be fair, human civilization has engaged in cognitive offloading for generations. We use calculators for complex arithmetic, GPS devices for navigation, and digital calendars to remember appointments because those tools free up cognitive space for higher-order tasks.
Reading, writing, and critical thinking, however, serve an entirely different purpose in an educational setting. The assignment is rarely just a finished product; the assignment is the practice. Every annotation, rough draft, revision, and classroom discussion strengthens the cognitive muscles that students rely on to analyze, reason, synthesize information, and make independent judgments.
Researchers are only beginning to examine the long-term consequences when generative AI assumes more of these traditionally human responsibilities. A study published by Microsoft Research and Carnegie Mellon University found that adult workers who relied heavily on generative AI reported less critical evaluation and independent reasoning during their work tasks. Although that specific study focused on adults rather than K-12 students, it raises profound implications for education. If adults become less likely to critically evaluate information when AI consistently performs the thinking for them, the potential impact on adolescents whose reasoning skills are still actively developing is a cause for serious concern.
Experienced educators note that they use AI nearly every day to brainstorm, challenge their own thinking, locate research, and consider perspectives they might have overlooked. Yet, AI never replaces their personal judgment. They still evaluate the evidence, make the final decisions, and write the final product. That distinction highlights the crucial difference between utilizing AI as a supportive learning partner versus allowing AI to become a complete learning replacement.
The Evidence Extends Beyond the Classroom
What individual teachers observe in their daily classes is corroborated by macro-level educational trends. According to the National Assessment of Educational Progress, only 30 percent of eighth graders performed at or above the “Proficient” level in reading, while one-third scored within the “Below Basic” category. Reading achievement remains stubbornly below pre-pandemic levels, meaning waves of students are entering high school with diminished comprehension, vocabulary, and analytical reading skills.
These metrics matter because reading comprehension underpins nearly every academic discipline. When students struggle to interpret complex texts, evaluate evidence, or construct meaning independently, the temptation to outsource that cognitive labor to an AI chatbot becomes nearly irresistible.
Higher education institutions are reporting remarkably similar concerns. University faculty increasingly describe incoming freshmen as less prepared to read complex texts independently, construct evidence-based arguments, and sustain analytical writing over the course of a semester.
A stark illustration of this trend occurred when a Brown University economics professor noticed that students had earned an average score of nearly 96 percent on a take-home midterm—far exceeding historical course averages. Suspecting widespread reliance on generative AI rather than genuine student understanding, the professor replaced the traditional take-home final with a proctored, in-person exam. The average score plummeted to under 49 percent. The experiment exposed an uncomfortable reality: polished work completed outside the controlled environment of a classroom does not always reflect what students can independently explain, analyze, or apply.
While that event occurred at the collegiate level, it mirrors the daily experiences of K-12 educators. It serves as a stark warning to school leaders that K-12 systems are ultimately responsible for ensuring students can succeed in postsecondary environments, whether that means college, career training programs, military service, or the modern workforce.
If students arrive at these postsecondary environments unable to think critically, explain their reasoning, or defend their own ideas, educators must engage in serious self-reflection. Schools must ask whether they properly prepared their students for the intellectual demands that await them. AI is not inherently responsible for this vulnerability; rather, the response to it lies with educators. This challenge invites teachers and administrators to intentionally design learning experiences that foster higher-order thinking, psychological resilience, and cognitive stamina.
In response, schools across the country are beginning to recalibrate their relationship with technology. Numerous school districts have expanded handwritten assignments, paper-based assessments, technology-free instructional blocks, and strict restrictions on personal mobile devices after educators reported tangible improvements in student attention, classroom discussion quality, and independent reasoning. In many communities, students are required to place their smartphones into locked pouches for the duration of the school day.
Ironically, while many schools have implemented phone-free policies to combat distractions, students often spend the remainder of the instructional day learning almost exclusively through laptops and tablets. The deeper issue may not be the technology itself, but the sheer volume of thinking that educators inadvertently ask students to surrender to it.
What This Looks Like in Practice
To combat these trends, educators are realizing the imperative of teaching students about artificial intelligence alongside digital citizenship, emphasizing that AI is meant to support learning rather than act as the primary thinker and creator.
In progressive classrooms, before students open a computer or touch an AI tool, they annotate texts independently, discuss ideas face-to-face with classmates, gather evidence, and develop an original claim. They wrestle with uncertainty because productive struggle is not an artificial barrier to learning; it is the fundamental mechanism through which true learning occurs. Only after students have established their own baseline thinking is AI invited into the instructional conversation.
Instead of asking AI to complete an entire assignment, students learn to use it as a sparring partner: generating counterarguments, identifying logical gaps in reasoning, suggesting additional textual evidence, or offering alternative interpretations. They compare the AI-generated responses with their own work, critique the strengths and limitations of each, and make final decisions independently.
Classrooms are also reintroducing practices that had been quietly abandoned, not out of nostalgia, but because they build cognitive endurance. Students complete handwritten brainstorming, annotate physical texts, participate in technology-free discussions, and periodically write entirely without digital devices. These are explicitly pro-thinking activities designed to establish healthy boundaries around technology, ensuring that screens do not dominate every waking minute of the school day.
Initially, this transition faces resistance. Students accustomed to immediate digital gratification often feel frustrated when required to plan arguments by hand before touching a keyboard. That discomfort, however, is not evidence that learning has halted; frequently, it is proof that deep cognitive work is finally taking place. Over time, students become more willing to defend their reasoning, ask deeper questions, revise thoughtfully, and trust their own intellect.
Moving Toward Cognitive Citizenship
Digital citizenship must evolve to encompass what some educators describe as cognitive citizenship: helping students understand precisely when technology strengthens their thinking and when it quietly replaces it. This requires teaching students to interrogate AI outputs rather than accepting them as infallible authority.
By encouraging students to question the underlying assumptions, biases, and missing perspectives in AI-generated text, educators shift learners from passive consumers into active evaluators. AI transitions from being an unquestioned oracle into just another perspective in the room.
Education ultimately exists to cultivate independent thinkers and develop growing minds capable of questioning, evaluating, and creating original answers. The future of learning does not require a false choice between traditional instruction and modern innovation, but rather a thoughtful balance that determines when each approach best serves the student.
While artificial intelligence can effortlessly generate an essay, it cannot wrestle with uncertainty, passionately defend an argument to its peers, or discover a profound idea through sheer human persistence. Those transformative moments remain uniquely human—and they are precisely the moments educators must continue to protect.