From the Desk of David Joyner
Return to the message archive.
Nov. 17, 2026
The fall term is beginning to wind down, and for me, that means I start looking toward the next semester. Enough of the semester has passed that I have a sense of what worked and what didn’t, and if I want to be ready for the next class to start, now is the time to start preparing the new semester’s syllabus.
For many of us, the last several semesters have been particularly challenging as rapid advancements in artificial intelligence (AI) have forced us to make more adjustments both between and within semesters. As we look toward a new term of AI-induced opportunities and challenges, here are a few questions faculty ought to consider about AI in the classroom, depending on what you’re trying to accomplish:
- If you’re using it to instruct students, ask: Are students using AI as a learning assistant or as a doing assistant? One of the traps of AI in learning is that it can blur the line between what the student knows and what they appear to know with aid. If the goal is to assess their learning, we need to make sure our assessments are consistently measuring what they can do alone — but if the goal is to teach them to do something, AI assistance may be a catalyst rather than a distraction.
- If you’re teaching students how to use AI, ask: What’s the common baseline of knowledge or familiarity you can assume? Are your students all relatively AI-savvy, or do they come from a wider spectrum of prior experience with the technology? Modern K-12 schools can vary from no AI allowed to no AI guardrails at all — can you find what your students’ baseline is and design accordingly?
- If you’re deciding what to teach in the first place, ask: Are you trying to target a capability gap that AI will soon close? Too much teacher time has already been wasted redesigning curriculum to focus on things AI isn’t yet capable of doing — if you’re only teaching it because AI can’t do it yet, it’s likely not a great skill to focus on. On the other hand, if you believe it’s worth learning even if AI can do it, then prioritize keeping it — just make sure to assess it in a more AI-resistant environment.
- If you’re writing a policy on student AI use, ask: How actionable, enforceable, and practicable is the policy really? Students are increasingly frustrated with vague AI policies that leave them guessing what is truly allowed, especially when it isn’t clear whether AI use could be conclusively identified. Any AI policy should be accompanied by an understanding of how violations will be conclusively demonstrated — and if a policy can’t be consistently enforced, the policy needs to change to be either more permissive or more restrictive, drawing lines around what we can know for sure.
- If students are completing an assignment using AI (either by design or because it can’t be prevented), ask: How are you visualizing and assessing the process instead of just the product? One of the chief challenges posed by AI is that the product is no longer a good proxy for the underlying process, so new checks must be introduced to let us assess and give feedback. In the best situations, this lets us address how students are using AI and provide guidance on how to use it more effectively.
- If you’re teaching students about AI, ask: Are you prepared to give your students an understanding of how to use AI responsibly? Will they come away able to engage in conversation about the psychological, environmental, and socio-political ramifications of the technology? AI has the potential to transform society in numerous ways, but to ensure it changes the world for good, we need people ready to have conversations about the wide range of potential impacts.
- If you’re using AI to generate your own content, ask: Do your students want the content to be correct, or for the content to be yours? Students know that AI has a wealth of knowledge about lots of topics, but often they’re not just looking for fundamental correctness — they’re looking for your view and your explanation. Even more, content is often just the bedrock on which a collegial relationship between teachers and students is formed: Outsourcing content production to AI may threaten that foundation if students are more interested in hearing from the real you.
- And no matter what you’re doing: Are you prepared to be transparent with your students about how you’re using AI and why? Transparency can feel risky: Opening up to our students about our reasoning can feel like opening our thought process up to challenges, but students are far more appreciative of a policy or approach where they can see the rationale than one handed to them as infallible dogma. And bringing students into the conversation about how we’re using AI helps them develop their own understanding of where it can be used effectively and where it can’t.
AI is moving too quickly for any sort of unified set of best practices to keep up. The only way we can ensure student success in this new era of learning is to collaborate on what works — with other educators and our students — and to be in conversation about what we’re experiencing.