Proven Practice, New Technology: What Marketing Educators Are Learning About AI in the Classroom

Many marketing educators share the same low-grade panic. Students have AI in their pockets, will use it whether we approve or not, and are entering industries already reorganizing around it. The question is no longer whether students will use AI, but how to teach them to use it well while keeping learning at the center. That question shaped the 2026 AMA Summer Academic Conference preconference for marketing educators, where a keynote, two classroom sessions, and an employer perspective converged on one lesson: AI can expand research, feedback, and practice, but students must remain responsible for verification, reflection, and judgment.

The Keynote: Seamless or Seamful?

The keynote, from Lorena Blasco-Arcas of ESCP Business School, framed the whole problem in a way that reset the room. Her argument: GenAI’s default mode is seamless, fluid, and frictionless, and that fluency creates false confidence. The fix isn’t less AI. It’s seamful design: putting friction back exactly where judgment is at risk of quietly migrating from student to machine.

She named three places to put it. Epistemic friction means students verify before use, locating and checking the AI’s sources themselves instead of accepting a polished answer. Procedural friction means instructors grade the reasoning trace rather than the output alone, so outlines, drafts, and revision history all count toward the grade. Dialogic friction means students must defend, dissent, and revise their work, not just submit it.

Her caveat matters as much as the framework itself: Friction isn’t neutral. It costs more for students with less support, so it must be scaffolded into an assignment, not simply imposed on everyone equally.

The Problem Isn’t Cheating. It’s Design.

Laurel Lane and April Schofield, from Metropolitan State University of Denver, faced three simultaneous goals: distinguish legitimate AI use from academic misconduct, preserve critical thinking, and prepare students for AI-enabled workplaces. A ban could not accomplish all three. Assignment redesign could.

Lane and Schofield built a repeatable process rather than a one-time fix. They pull input from outside the classroom, drawing on the insights of program partners and advisory boards who can say what “AI-ready” looks like on the job. They identify where AI already shows up in the discipline’s real workflow. They design assignments that require both AI use and a human judgment step on top of it. And they have students explain and defend what they changed, kept, or rejected from the AI output. That last step protects the learning: If a student can’t articulate what they did differently than the AI, the assignment isn’t doing its job yet.

If a student can’t articulate what they did differently than the AI, the assignment isn’t doing its job yet.

Scaling Practice, Not Just Assigning It

Amy Vandaveer Novak from the University of Houston’s C. T. Bauer College of Business took the same principle further and scaled it. Educators in their professional selling program tackled a different version of the same problem: How do you give 1,350 students the live-practice experience that used to require an instructor and a role-play partner in the room? The answer was AI-mediated role-play simulation, layered on top of the traditional sales curriculum rather than instead of it. Students still learned SPIN selling and objection handling the traditional way: AI just gave them somewhere to practice at scale, with immediate feedback and no scheduling bottleneck.

Repetition without reflection doesn’t move the needle. AI can deliver unlimited reps; it can’t make a student reflect. That’s still the instructor’s job: to design for the same dialogic and procedural friction the keynote called for, just running at scale.

The Hiring Side: What Actually Gets You Hired

The third session came from outside the classroom. Kate Klein, executive vice president of marketing at Houston Fitness Partners, a 53-location Planet Fitness franchisee, walked the room through what “ready” looks like from her side of the hiring table. Her framing: It isn’t a credential test; it’s a judgment test of whether this person can move the business today. School teaches frameworks, channels, and campaign mechanics. The job demands judgment under ambiguity, ownership of a number, and the ability to connect an activity to a business outcome. The gap between the two is exactly where hiring decisions get made.

Klein was specific about what closes that gap. Candidates who get a second look are those who can walk through one campaign end to end (spend, audience, outcome, and what they’d change) without needing the question softened. They say, “This underperformed,” and explain why, instead of only celebrating what worked. They default to a point of view and can defend it with data. And increasingly, they understand AI tools well enough to direct them, not just use them.

That last point reinforces what the keynote and both teaching sessions emphasized. A student who can explain, defend, and revise their AI-assisted work isn’t just protecting academic integrity; they’re practicing the exact accountability Klein screens for on the other side of the hiring table.

Five Practical Starting Points

If you are rethinking an assignment for your next course, consider five practices that emerged across the sessions:

1. Define AI’s job. Give the tool one narrow role, such as research assistant, brainstorming partner, role-play partner, or first-draft generator. Clear boundaries make it easier for students to understand what AI should and should not contribute.
2. Preserve the student’s job. Require students to interpret, select, apply, or defend the output. AI may generate possibilities, but students should remain responsible for the decisions that shape the final work.
3. Build in verification and reflection. Ask students to check factual claims against credible sources and explain what they accepted, rejected, or changed. A short defense of those decisions protects learning more effectively than relying on an AI-detection tool.
4. Use AI to expand practice, not bypass it. The strongest applications give students more opportunities to practice skills they are already expected to learn. AI can provide additional repetitions and faster feedback, but repetition matters only when students apply what the feedback teaches them.
5. Scaffold the friction. Verification, documentation, and defense requirements should come with examples, source-evaluation guidance, and transparent grading criteria. Productive friction should strengthen learning, not create an additional barrier for students who have less experience or support.

Proven practice and new technology are not competing choices. The strongest classrooms use AI to expand opportunities for research, feedback, and repetition while keeping students responsible for judgment. That shift changes how instructors write assignments, how they assess learning, and how confidently students can explain their work when they enter the workplace.

The authors of this article co-chaired the marketing educators’ preconference at the 2026 AMA Summer Academic Conference.

Read more from AMA Career & Classroom