AI Undermines Reading Habits, Threatening Literacy & Critical Thinking
A perfect storm is brewing for the act of reading. As both children and adults already dedicate less time to books than in the recent past, artificial intelligence has emerged, dramatically reshaping how individuals engage with written material. From assigned texts to research papers or leisure reading, AI’s influence is profound, raising concerns that it is accelerating a shift in the perceived value of reading as a fundamental human endeavor.
While AI’s prowess in generating text has garnered significant attention, its sophisticated ability to “read” vast datasets and subsequently produce summaries, analyses, or comparisons of books, essays, and articles is only now becoming a focal point for researchers and educators. The convenience of obtaining an AI-generated plot summary or thematic analysis for a novel assigned in class, for instance, risks undermining students’ motivation to engage with the full text. This phenomenon, while amplified by AI, isn’t entirely new. Tools like CliffsNotes have provided text summaries since the late 1950s, and scholarly abstracts became ubiquitous in the mid-20th century. The internet further expanded these shortcuts, with services such as Blinkist condensing non-fiction books into 15-minute audio or text summaries.
However, generative AI elevates these workarounds to an unprecedented level. AI-powered applications like BooksAI offer summaries and analyses once exclusively crafted by humans, while platforms such as BookAI.chat invite users to “chat” directly with books, eliminating the need to read them at all. Consider a student tasked with comparing Mark Twain’s The Adventures of Huckleberry Finn with J. D. Salinger’s The Catcher in the Rye. While CliffsNotes might provide individual summaries, the comparative analysis traditionally required human effort. Now, advanced large language models or specialized tools like Google NotebookLM can handle both the “reading” and the comparison, even generating insightful questions for classroom discussion. Yet, this efficiency comes at a cost: the profound personal growth derived from vicariously experiencing a protagonist’s struggles, a core benefit of engaging deeply with such literature, is lost.
In academic research, AI tools such as SciSpace, Elicit, and Consensus combine search engine capabilities with large language models to locate, summarize, and synthesize relevant articles, drastically reducing the time needed for literature reviews. Publishers like Elsevier even promote their ScienceDirect AI, proclaiming “Goodbye wasted reading time. Hello relevance.” But this streamlined process bypasses a crucial intellectual step: the independent judgment of relevance and the spontaneous forging of connections between disparate ideas.
These advancements arrive amidst a discernible decline in reading habits, a trend that predates the widespread adoption of generative AI. In the United States, the National Assessment of Educational Progress reported a significant drop in fourth graders who read for fun almost daily, from 53% in 1984 to just 39% in 2022. For eighth graders, the decline was even starker, plummeting from 35% in 1984 to 14% in 2023. Across the Atlantic, the U.K.’s 2024 National Literacy Trust survey found that only one in three 8- to 18-year-olds enjoyed reading in their spare time, nearly a nine-percentage-point decrease from the previous year. Similar trends are evident among older students; a 2018 survey of 600,000 15-year-olds across 79 countries revealed that 49% read only when required, up from 36% a decade earlier.
The situation in higher education is equally concerning, with numerous reports chronicling the diminishing amount of reading undertaken by American college students. Research indicates that university faculty are assigning less reading, often in response to students’ reluctance to engage. Cultural commentator David Brooks highlighted this issue poignantly, recalling how a group of graduating students from a prestigious university struggled to name a single book that had profoundly impacted them, admitting they only “sampled enough of each book to get through the class.”
Adult reading habits paint a similarly bleak picture. According to YouGov, only 54% of Americans read at least one book in 2023. South Korea’s figures are even lower, with just 43% of adults reporting reading a book in 2023, a dramatic fall from nearly 87% in 1994. The Reading Agency in the U.K. observed similar declines, noting that 35% of adults in 2024 identified as “lapsed readers”—individuals who once read regularly but no longer do, with 26% citing time spent on social media as a primary reason. This term now broadly encompasses anyone who de-prioritizes reading, whether due to lack of interest, increased social media consumption, or increasingly, reliance on AI to do the reading for them.
The justifications for reading are myriad: pleasure, stress reduction, learning, and personal development. Studies correlate reading with enhanced brain growth in children, increased happiness, longevity, and a slowing of cognitive decline. This last point is particularly pertinent given the growing tendency for individuals to delegate cognitive tasks to AI, a phenomenon known as “cognitive offloading.” Emerging research demonstrates the extent to which people engage in cognitive offloading when using AI, revealing that the more users rely on AI for tasks, the less they perceive themselves as drawing upon their own intellectual capacities. One study using EEG measurements found distinct brain connectivity patterns when participants used AI to assist in essay writing compared to writing on their own.
While it is premature to fully understand AI’s long-term impact on our independent thinking abilities, current research primarily focuses on writing tasks or general AI usage, not specifically reading. However, if the practice of reading, analyzing, and formulating personal interpretations diminishes, these vital cognitive skills are at risk of weakening. Beyond intellectual faculties, relying heavily on AI for reading also deprives individuals of the intrinsic joys of the activity—the delight of encountering a poignant dialogue, savoring a finely crafted phrase, or forging a deep connection with a character. AI’s promise of efficiency is undoubtedly tempting, but it carries the significant risk of undermining the very benefits of literacy itself.