It’s never been easy for students, at either the college level or younger, to read and analyze challenging texts. In the past, many students sought a summary from Wikipedia, CliffsNotes, or an encyclopedia.
The rise of artificial intelligence and large language models (LLMs) introduces a new wrinkle in the search for understanding what we read, with lightning-fast AI-produced summaries, outlines, or deeper analysis.
At the same time, people are reading less and less. Nearly half of Americans read zero books in 2023, according to a YouGov survey. The steep decline is particularly evident among college graduates.
Children also exhibit a decline in literacy proficiency. National Assessment of Educational Progress data shows 41 percent of fourth graders and 34 percent of eighth graders are below basic reading levels as of 2024.
People read much less for pleasure too, notes Margaret Vaughn, a literacy expert and professor in the College of Education, Sport, and Human Sciences at Washington State University. “Among eighth grade students, 83 percent report that they do not frequently read for enjoyment,” she says.
While not all the literacy struggles are due to increased AI use, it contributes to how colleges and schools adapt ways of teaching how to read. It also leads to speculation about a “post-literate culture” in which media are consumed through screens rather than denser texts.
The AI era causes concern, and shows some promise, for WSU faculty who teach reading. “We’re seeing a really radical shift in this moment. If I’m a student and that text is too complicated, there’s an easy route now that wasn’t there just two or three years ago,” says Kate Watts, a professor in the WSU Department of English.
The risk with LLMs is that students may not develop critical thought, Watts says. She explains that AI-generated essays often lack true analysis and synthesis, and typically produce an amalgamation rather than novel thinking.
Rachel Sanchez (’09 MA English), a professor of English and WSU’s director of composition, agrees. “I don’t want them to offload their critical thinking skills to something else they’re not in control of,” she says.
The use of AI and LLM gets complicated by external factors that are often overlooked. When Sanchez assigned an article lamenting the lack of student reading, her class offered a key insight. “Students explained they were busy, balancing jobs, managing family obligations, and struggling to find the time to commit to complex material,” she says.
Practical pedagogical changes and adaptive strategies, however, can meet students where they are, Sanchez says.
“I’m not sure as educators that we’re adequately explaining why it’s important to grapple with a complex text,” Watts says. “The magic actually happens in reflection. It happens in asking them to think about that experience, and what does that mean for the next time?”
To ensure students develop genuine critical skills, WSU instructors in English advocate for teaching reading in a strategic way. Watts advises instructors to teach intentional reading methods, such as fast-forwarding to a text’s conclusion and working backward to analyze necessary data points.
Learning how to use AI effectively also requires opportunities for failure and reflection. Watts describes a task in which students use an AI tool to summarize an article and then reflect on the result, comparing it with their neighbor’s summary. Students determine the source of any discrepancy.
“Was it a flaw in the AI, an issue with their prompt, or a gap in their own comprehension? This moves students past a simplistic ‘good or bad’ understanding of AI to recognizing that the technology is complicated and requires informed critical engagement,” Watts says.
Teachers at the K-12 level also need to shift their methods in reading instruction, and they need more support for adaptation, so their students engage more with reading, Vaughn says.
Vaughn, who also hosts a literacy podcast called Getting Smarter, advocates for teachers to be allowed more flexibility to develop a love of reading in students.
“It’s much easier to follow a script than to be flexible and adaptive in the moment,” she says. “But we have to shift our view about teachers. Teachers are skilled professionals. They didn’t go into the profession to read a script.”
Vaughn shares an example of an elementary teacher reading about volcanoes. “When the teacher noticed students passionately sharing their personal experiences hiking and exploring mountains like Mount St. Helens, she stopped her planned lesson,” Vaughn says. The teacher pivoted, and the students’ spontaneous excitement led to two and a half weeks of extensive, authentic research driven by student questions.
Vaughn also recommends a “tackle” approach to get kids connected to reading. Parents and teachers can follow a few simple principles: time to read; access to high interest, authentic, and culturally responsive text; choice in reading material; and having students engage with books and with each other.
AI can support adaptive strategies, Vaughn says, such as providing assistance for students who struggle or have disabilities. It can also generate content that students then analyze and edit for errors.
The goal is not to eliminate technologies like AI and LLMs, but to manage them critically. Whether it’s for K–12 or an English 101 class at WSU, students can gain confidence in reading and analysis, even when powerful tools promise to do the thinking for them.
Learn more
Getting Smarter podcast with Margaret Vaughn
Why is reading for fun plummeting in the US? Plus: tips for families (December 31, 2025, University of Miami)