This week’s reading news is not really about whether screens are good or bad. It is about whether students are still doing the slow, effortful work that deep comprehension requires when the text arrives through a laptop, tablet, phone, e-reader, or AI summary.
The timely signal comes from the new PISA 2025 Results from the OECD, released on September 8, 2026. More than 760,000 15-year-olds in 91 countries and economies took part. The OECD reported that average performance across member countries fell to the lowest level PISA has recorded in science, reading, and mathematics, with reading and mathematics drawing particular concern. The report also examines education in the age of AI, misinformation, and digital learning environments.
OECD’s follow-up discussion, PISA Insights: Falling scores, AI concerns and digital distractions, sharpens the practical question. AI use is not one simple variable. Students who use AI for specific schoolwork tasks such as summarising texts or drafting can perform worse than non-users, while students with opportunities to learn AI literacy and use AI more generally can show different outcomes. The pattern fits a broader learning-science warning: tools help when they extend thinking, but they hurt when they replace it.
That warning matters for reading on screens because digital text often rewards speed. A student can scroll, search, highlight, copy, summarize, and ask a chatbot before forming a durable mental model of the text. The session may feel productive because the page was finished and the main points were visible. But deep comprehension is tested after the page is gone: can the learner explain the author’s claim, evidence, structure, limits, and implications from memory?
Recent research gives educators a reason to be careful. Teachers College, Columbia University summarized brain-imaging evidence in Children Derive Deeper Meaning from Printed Texts Than Screens, reporting that children reading print showed stronger engagement with deeper semantic processing than children reading from screens. The finding does not mean every screen reading task is weak. It does suggest that younger readers may need more support when the goal is meaning, inference, and integration rather than quick information pickup.
Another useful source is the International Association for the Evaluation of Educational Achievement’s open teacher resource, Teaching Reading Comprehension in a Digital World. Its central implication is balanced: digital reading is now a real literacy context, so learners need explicit strategies for navigating online texts, evaluating information, and maintaining comprehension in digital environments. The answer is not to pretend screens are leaving school. The answer is to teach students how to read well when screens are part of the task.
AI adds a newer layer. In the 2026 preprint How College Students Use AI to Navigate Course Readings, researchers followed undergraduates for eight weeks and analyzed how they used AI during assigned readings. Many prompts were about comprehension, but fewer were metacognitive prompts that asked students to monitor their own understanding. The researchers described a risk of students reading through AI rather than with the text, using AI output to decide which parts deserved attention.
What Happened
The news is that international assessment data, digital-reading research, and AI-reading studies are converging on the same concern. Students are surrounded by tools that can make reading faster, more searchable, and easier to outsource. At the same time, the skills needed for deep comprehension – sustained attention, self-questioning, source evaluation, inference, and recall – are becoming more important, not less.
The evidence does not support a simple ban on digital reading. Screens can provide accessibility features, dictionaries, translation, annotation, search, multimedia context, and immediate links to background knowledge. But those benefits do not automatically produce understanding. A screen becomes a learning tool only when the learner uses it to think more clearly, not merely to move through text more quickly.
Why It Matters
Deep comprehension is slower than recognition. A learner can recognize a sentence while rereading it, agree with a summary, and still fail to reconstruct the argument later. That is why screen reading can create a dangerous kind of fluency: the material feels available because it is open, searchable, highlighted, or summarized.
For teachers, tutors, parents, and adult learners, the practical issue is task fit. A short announcement, low-stakes search, vocabulary lookup, or accessible text-to-speech reading may fit a screen perfectly. A dense chapter, difficult article, primary source, legal case, lab reading, or exam-preparation passage may require print, a larger screen, a stable page view, or a deliberately slower digital setup.
The Practical Learning Conclusion
Use the reading format that matches the learning goal, then build in checks that prove understanding before moving on.
- Choose print or a stable page view for hard reading. If the goal is deep comprehension, avoid narrow phone screens and endless scrolling when another format is available.
- Pause before the next scroll. At the end of each section, state the main claim in one sentence before continuing.
- Annotate for thinking, not decoration. Replace broad highlighting with short notes such as “claim,” “evidence,” “confusing,” “example,” or “question.”
- Use AI after your first attempt. Explain the passage yourself, then use a chatbot or summary to check gaps, generate questions, or compare interpretations.
- Close the text and retrieve. Write the main point, two supporting details, and one implication from memory. Reopen the text only after trying.
- Switch formats when the task changes. Use screens for search, accessibility, quick review, and interactive work. Use print or slower screen settings when the task demands sustained reasoning.
- Check transfer. Apply the reading to a new example, problem, debate, or decision. If the idea cannot travel, comprehension is still shallow.
The conclusion is simple: digital reading needs deliberate friction. When learners slow down, check their own understanding, and decide when print or screen fits the task, screens can support learning. When the device turns reading into scrolling, searching, and outsourcing, comprehension becomes fragile. The deeper lesson from the latest evidence is to protect the thinking that happens between the sentences.