From Social Media to AI
Artificial intelligence is increasingly being blamed for declining literacy among students. There are legitimate reasons for concern.Students can now ask AI to summarize a text, correct their writing, construct an argument or even produce an entire assignment.However, focusing exclusively on AI risks overlooking a much older influence on students’ literacy: social media.
The Literacy Environment Has Already Changed
Long before generative AI entered classrooms, social media had fundamentally changed the way young people encountered written language.
Much everyday digital communication consists of fragments rather than complete sentences. Punctuation disappears, capitalization becomes optional, emojis replace words and sometimes entire responses. Abbreviations such as LOL, IDK, TBH, IMO, POV, NGL and IRL communicate meaning quickly without requiring conventional sentence construction.
The problem is not that young people use slang or abbreviations. Every generation has done that. The concern is the sheer amount of time students spend in a language environment dominated by short, fragmented communication. Social media also encourages rapid consumption: scroll, scan, watch, react. Move on.
Academic literacy requires something very different. Students need to sustain attention across paragraphs and pages, follow an argument, interpret vocabulary in context, recognize relationships between ideas, distinguish important information from supporting detail and draw conclusions from what they have read. In other words, they must comprehend.
Then Came AI
A recent discussion with Dr Mickie de Wet, researcher in AI ethics and governance and business innovation at Católica Lisbon School of Business and Economics, raised an interesting question: Is AI necessarily damaging to literacy, or could the way students interact with AI actually provide opportunities to strengthen it?
AI communicates very differently from social media
Ask an AI system a question and it generally answers in complete sentences. It uses conventional spelling, punctuation and grammar. It organizes information into paragraphs, explains concepts and provides context. Interestingly, it normally does this even when the person communicating with it does not.
For example, a student might type: “why romeo kill himself he didnt know juliet wasnt dead?”
AI can understand the intended question, but its response is likely to explain Romeo’s actions using complete, grammatically conventional sentences. The student has entered the conversation using weak written English but is receiving a model of considerably better written language. More importantly, the student has to read the answer. Reading is not enough. The student has to comprehend it.
AI as a “Glorified Search Engine”
Teachers sometimes complain that students are using AI as little more than a glorified search engine. That may be true, but even this deserves closer consideration. A conventional search engine such as Google, may provide students with pages of links. AI is more likely to provide several paragraphs of explanation. Students can then ask for clarification, request a simpler explanation, investigate an unfamiliar term or ask a follow-up question. The interaction becomes a written conversation. This does not automatically make it educational. AI can be wrong. It can misunderstand information, invent references, confidently present inaccurate information and reproduce poorly researched material.
This is precisely where another important literacy skill enters the picture:
Literacy Now Includes Verification
Reading comprehension has traditionally required students to understand what a text says. In an AI environment, students must go further. They need to ask:
Is this accurate?
Where did this information come from?
Does the source actually support the claim?
What information has been omitted?
Is there another interpretation?
Can I verify this elsewhere?
These are not lesser literacy skills. They are sophisticated forms of critical literacy. The ability to read an apparently authoritative response and question its accuracy may become one of the most important literacy skills students develop.
AI and Students with Weak Written English
AI also presents an interesting possibility for students whose written English is weak. If a student communicates using poor grammar, confused sentence construction or badly misspelled words, AI can usually determine the intended meaning and respond using conventional written language. Students are therefore repeatedly exposed to models of clearer English.
AI could take this one step further by occasionally responding, “I understand what you are asking. A clearer way to write your question would be…”
It could then model the improved sentence before answering the question. This should not happen constantly. Continuous correction would quickly turn a useful conversation into an irritating grammar exercise. Used selectively, however, it could help students recognize the difference between the language they produce and clearer, more accurate written communication.
The Danger Is Still Real
None of this means that AI is automatically good for literacy.
If students ask AI to write an essay and submit it without reading or understanding it, no literacy development has taken place.
If AI continually constructs their arguments, chooses their vocabulary, organizes their paragraphs and corrects every sentence, students may gradually lose opportunities to develop those abilities themselves.
The important distinction is,who is doing the thinking?
AI can replace the literacy task, or it can become part of the literacy task.
That distinction should be central to the way schools develop policies and professional development around classroom AI use.
What Can Teachers Do?
Instead of concentrating exclusively on preventing students from using AI, teachers can design activities that require students to think about what AI produces.
Students can:
compare an AI summary with the original article and identify what was omitted
investigate an AI response and identify inaccurate or unsupported information
compare a poorly written prompt with an improved version and examine how the response changes
challenge an AI interpretation of a text using evidence from the original
verify references supplied by AI
compare the language used in social media, AI responses and formal academic writing.
In each case, AI is not doing the literacy work for the student. AI becomes the material on which the student performs the literacy work.
A Professional Development Question
This also has implications for teacher professional development. Simply telling teachers that students should or should not use AI is unlikely to be sufficient. Teachers need opportunities to explore how AI can be incorporated into lessons without removing the thinking, reading, writing and reasoning that the lesson was designed to develop.
PLCs can examine student AI use in much the same way they examine other aspects of learning:
What are the students’ learning needs?
What literacy skills are we trying to develop?
How could AI support those skills?
At what point would AI begin doing the work the student needs to learn to do independently?
These questions move the discussion away from whether AI is simply “good” or “bad” and towards its effect on student learning.
Perhaps We Are Asking the Wrong Question
Technology has unquestionably changed literacy. Social media created an environment dominated by speed, fragments, abbreviations and rapidly changing content. AI has introduced a different challenge. It can produce polished, conventional language almost instantly. However, polished language is not necessarily accurate language. Students therefore need more literacy education, not less. They need to read carefully enough to understand an AI response, think critically enough to question it and possess sufficient knowledge to recognize when something does not make sense.
How do we stop students from using AI, should perhaps no longer be the educational challenge. A better question might be,how do we teach students to use AI without allowing it to do the thinking for them?
Perhaps the greatest danger to literacy is not that students will use AI. It is that schools will fail to teach them how to use it well.