AI in the Classroom: 11 Realities Instructors Are Facing

AI in education is no longer a future-tense conversation; it is already shaping how students brainstorm, write, and study. That leaves instructors working through the harder part, which is what belongs in a classroom, what needs clearer boundaries, and what has to be redesigned altogether now that the rules have changed.

According to the Tyton Partners “Time for Class” 2026 Report, 47% of faculty and 61% of administrators report that their roles have fundamentally changed, or will within three years as instructors work through how to best support students.

The same Tyton Partners Report noted that assessment is the number one teaching practice faculty say they are modifying because of AI. Given that faculty concerns about student cheating have climbed from 36% in 2024 to 55% in 2026, this means we are looking at a story that is less about technology and more about teaching.

We borrowed 11 ideas from The What & Who of EDU’s most recent podcast “AI in the Classroom: What’s Working, What’s Not, and What Instructors Are Still Figuring Out” to give you the quick and dirty tips. But be sure to check out the podcast on Apple, Spotify or YouTube if you’d like the full scoop.

1. Stop Playing Pretend: AI Doesn’t Just Go Away

Ignoring the storm won't stop the rain. According to Dr. Ryan Herzog, Associate Professor of Economics, Program Coordinator, and Faculty Fellow at Gonzaga University, pretending AI doesn't exist leads to students using it as a Magic 8 Ball, shaking it for an answer without understanding the underlying logic. The goal should be guided use; helping students to explore problems, test ideas and think through processes instead of grabbing a polished answer and calling it good.

2. Keep AI in the "Messy Middle"

AI works best in the draft stage, not the finish line. By using AI specifically for brainstorming, refining ideas, or experimental processes, the "messy middle" reinforces the importance of the writing and thinking process over the final product. Dr. Cristin Monroe, Educational Research Associate at the Center for Innovation in Teaching and Learning at the University of Illinois reminds us that even students who feel skeptical about AI’s environmental or ethical impact may still find it useful as support.

3. Don’t Make AI a Secret

According to Jennifer Duncan, Associate Professor of English at Georgia State University's Perimeter College, when instructors treat AI as a forbidden secret, they may inadvertently create an atmosphere of dishonesty. Students will still use it, just quietly. Being open about AI, even showing vulnerability as a learner, humanizes the course concepts and establishes ethical boundaries.

4. Pick a Lane: Just Not One of the Extreme Ones

Banning AI entirely and giving it a free pass are both easy answers. There is another path, which is to make more intentional use of it. According to Betsy Langness, Psychology Department Head at Jefferson Community and Technical College, Psychology offers a perfect gateway to discuss ethics, critical thinking, and self-awareness. AI is neither inherently good or bad, but how we use it to support student learning has an impact.

5. Don’t Turn It Into a "Gotcha" Machine

The minute AI becomes only a detection problem, the teaching part starts to shrink. While faculty concerns about cheating have jumped from 36% to 55% since 2024, focusing entirely on surveillance turns the instructor into a cop rather than a mentor. Dr. Sara Lahman, Professor of Biology and STEM Outreach Coordinator at the University of Mount Olive reminds us that students need guidance on prompts, creativity, and process, not just suspicion about whether they used a tool.

6. Swap Strategies: Community is the Answer

Dr. Star Sinclair, Assistant Professor of Psychology at Florida Gulf Coast University enthusiastically reminds us that instructors should not have to figure this out alone. Some of the most useful AI support on campus may not be a tool at all. It may be another instructor who has already tried something, learned from it, and is willing to share. Faculty conversations like that make the whole thing feel less abstract and a lot more usable.

7. Use It

Dr. Daniel Look, Charles A. Dana Professor of Mathematics at St. Lawrence University knows that calculators didn't kill math; they evolved it. He thinks AI is not so different and believes we must integrate AI as a tool for brainstorming and assistance without letting it replace human judgment. AI can think fast, but students still need to learn to think clearly, knowing how to distinguish between a breakthrough idea and AI slop.

8. Clarify what that "A" Actually Actually Means

Here’s the existential question keeping everyone up at night: when AI can do the heavy lifting, or when smart glasses might whisper the right answer during an exam, what does a grade actually signal anymore? Dr. Erika Martinez, Professor of Instruction at the University of South Florida, reminds us that we’re balancing two jobs: teaching students so they gain knowledge, and certifying to the world that they’re actually capable. We’re all scrambling to figure out how to guarantee proficiency when our old safety nets, like in-class assessments, are facing such a massive tech upgrade

9. Don’t Remove the Struggle Without Replacing the Skill

Dr. Martinez also reminds us that when you remove the struggle, you may accidentally delete the skill. Just as looking up a word in a physical dictionary taught students search, self-correction, and task isolation, AI removes that hard part. If we take away one kind of struggle, we may need to build the underlying skill back in somewhere else.

10. Keep the Hard Part Hard

AI can make it much easier to generate large banks of quiz questions. But Jennifer Ripley Stueckle, Teaching Professor and Non-Majors Biology Program Director at West Virginia University, reminds us that ease is not the same thing as rigor. When assessments drift toward quick knowledge checks, higher-order skills like analysis, evaluation, and application can quietly slip out of the course. The hard questions are usually hard for a reason.

11. Redesign for Authentic Assessment

Even if the goals are the same, assignments that worked perfectly in 2022 might now feel obsolete. That just means the design needs another look. When students move from pitch to draft to revision, or build an argument in ways the instructor can actually see, the learning becomes harder to fake and easier to trust, according to Kendra Thomas, Associate Professor of Psychology at Hope College.

What this conversation keeps surfacing is how much pressure AI is putting on parts of teaching we used to leave unexamined, like the assignments that used to work, the grades that felt self-explanatory, and the fine line between help and replacement.

However many teachers are positively embracing this change, learning how to work alongside AI to develop entirely new processes. Their day-to-day experience is providing meaningful insights into how to best leverage these tools to achieve results that actually benefit their students.

The instructors in this episode are not offering a single rule for every classroom. Instead, they are demonstrating what it looks like to respond with clarity: clearer expectations, better assessment design, and more visible learning. That may be the real job right now.

Pick one assignment and look at it again. That is usually where the useful work starts.

FAQs

How are instructors handling AI tools right now?

Instructors are moving away from traditional take-home essays. They are using scaffolded assignments, in-class writing sessions, and interactive digital learning platforms that track student progress over time.

How can I design assignments that resist AI cheating?

Focus on highly specific prompts that ask students to connect course concepts to their personal experiences or specific, recent local events. You can also grade the individual steps of a project rather than just the final product.

Do AI detectors actually work for college essays?

Most studies show that AI detectors have high rates of false positives and can be bypassed easily by changing a few words. Relying on them often hurts student trust without solving the underlying academic integrity issue.

What is Achieve?

Achieve delivers research-backed personalised learning, real-time feedback, and accessibility features—all integrated into your LMS to keep students engaged without increasing your workload.

Further Reading

Navigating AI's Impact on Teaching and Learning

5 Questions Science Educators Should be Asking About AI

Cheaters Never Win, But Are They Really Cheating? Rethinking Integrity in the AI Classroom