Universities Are Making AI Literacy Mandatory — But Teachers Say It's Killing Critical Thinking

2026-06-14 · Nia

There's a fascinating contradiction playing out in higher education right now. Universities are racing to make AI literacy a graduation requirement — treating it as essential as writing or math. At the same time, teachers are loudly warning that AI is destroying the very cognitive skills that made those other competencies worth teaching.

Both sides are right. And that's what makes this moment so interesting.

SUNY Draws the Line

On April 30, 2026, the State University of New York's Board of Trustees approved a system-wide AI policy that will reshape how 64 campuses approach artificial intelligence. Starting this Fall, AI literacy becomes an official component of the Information Literacy core competency for all incoming undergraduates.

This isn't a vague suggestion to "incorporate AI." SUNY's policy requires courses to embed bias detection, prompt evaluation, and ethical reflection into the curriculum. Campuses must classify AI tools into risk categories — minimal, moderate, and high — with higher-risk systems requiring impact assessments and human oversight. Every campus needs a compliant AI policy published by December 31, 2026.

It's the most concrete institutional commitment to AI education we've seen from a major U.S. public university system. And SUNY isn't alone — China and the UAE have mandated AI education starting in the 2025-26 school year. The message is clear: AI fluency is no longer optional.

The Teacher Revolt

Here's where it gets complicated.

An NPR/Ipsos poll released June 5, 2026 found that 54% of K-12 teachers believe AI is making it harder for students to develop critical thinking skills. That's not a fringe opinion — it's the majority view among the people who see students interact with these tools every day.

The numbers paint a grim picture: 55% of educators say students primarily use AI as a shortcut to avoid work. 59% say AI tools are eroding trust between teachers and students. 57% report the technology hinders their ability to assess what students actually know.

And yet — here's the twist — 60% of those same teachers use AI themselves for lesson planning, material development, and administrative tasks, saving an average of nearly 6 hours per week.

Teachers aren't anti-AI. They're anti-lazy-AI-use. There's a meaningful difference.

The 95% Problem

According to Coursera's 2026 higher education guidance report, 95% of students and educators are already using AI tools. Among students, 94% use generative AI for assessed work. But only 26% of institutions have formal guidelines governing that use.

Read that again: nearly everyone is using AI, and nearly nobody has rules for it.

This is the policy vacuum that mandates like SUNY's are trying to fill. But there's a deeper question: are universities teaching students to use AI, or teaching them to think despite AI?

The Stanford AI Index 2026 Report offers a hint at where things are heading. New AI PhDs in the U.S. and Canada grew 22% from 2022 to 2024, with — surprisingly — most of that growth directed toward academia rather than industry. The proportion entering academia rose to nearly 32% in 2024, reversing a decade-long brain drain to Big Tech.

Universities aren't just teaching AI. They're reclaiming it.

What Good AI Literacy Actually Looks Like

The mistake most institutions make is treating AI literacy as "how to use ChatGPT." That's like teaching someone to Google without teaching them to evaluate sources. It's technical competency without intellectual depth.

SUNY's approach gets closer to the mark. Their framework incorporates three layers:

Technical fluency — Understanding what models can and can't do, how prompts shape outputs, what hallucination means

Critical evaluation — Detecting bias, assessing source reliability, understanding when AI outputs are wrong (or worse, plausibly wrong)

Ethical reasoning — Grappling with questions of attribution, academic integrity, and the societal implications of automated decision-making

Syracuse University's new AI and Emerging Media minor takes a similar approach, combining generative tool training with data strategy and media ethics. UConn's "AI for ImpaCT" initiative spans education, research, public engagement, and workforce development.

These programs share a common thread: they don't just teach AI as a tool. They teach it as a lens for understanding how knowledge is created, validated, and distributed in the modern world.

The Real Crisis Isn't AI — It's Passivity

Here's my take, and it might be unpopular: AI isn't destroying critical thinking. It's revealing how fragile our teaching of critical thinking always was.

If a student can submit an AI-generated essay and get an A, the problem isn't the AI — it's the assessment. If students treat every tool as a shortcut, that's a mindset problem that predates ChatGPT by decades. Remember CliffsNotes? SparkNotes? Wikipedia copy-paste essays?

What AI does is make intellectual passivity easier at the exact moment we need intellectual engagement the most. The students who learn to use AI as an amplifier — for research, for iteration, for exploring perspectives they wouldn't encounter otherwise — will outperform everyone. The students who use it as a crutch will produce mediocre work that looks polished.

The real mandate isn't teaching students to use AI. It's teaching them to think harder because AI exists.

As we discussed in AI literacy is the new core competency every student needs, this isn't just about university requirements — it's about fundamental preparation for the modern workforce. And the gap between students using AI and understanding it is widening faster than policies can close it.

What Builders Should Watch

For anyone building in the ed-tech space, the signals are clear:

Assessment tools are the next battleground. Universities desperately need better ways to evaluate learning outcomes in an AI-enabled world. The current plagiarism-detection approach is a dead end.
AI literacy platforms will be mandatory purchases. SUNY's 64 campuses alone represent massive demand for structured AI literacy curricula.
The 60% teacher adoption rate is an entry point. Teachers are already using AI — they need better tools designed for their specific workflows, not repurposed consumer products.

The Gallup/Walton Family Foundation survey found that over 60% of teachers using AI lack formal guidance. That's not a failure — it's a market opportunity for anyone who can deliver structured, educator-specific AI training at scale.

The Bottom Line

Universities making AI literacy mandatory is the right call. But the real work isn't in the mandate — it's in the execution. Teaching students to prompt an LLM is trivial. Teaching them to question the output, evaluate its sources, recognize its biases, and decide when to ignore it entirely? That's education.

SUNY's policy is a starting gun, not a finish line. The institutions that get this right won't just produce better AI users — they'll produce better thinkers. And in a world increasingly saturated with machine-generated content, better thinkers are exactly what we need.

The universities treating AI as infrastructure rather than experiment are the ones to watch. They're building for the long game while everyone else is still debating the syllabus.