Embracing AI A Guide for Humanities Educators

By Sarah Tangeman

Welp. It’s happened.

AI is no longer a futuristic concept knocking at the door of higher education—it’s already here. For those of us teaching in the humanities, especially fields like my area of focus, Art History, it can be difficult to envision what responsible, thoughtful AI integration looks like. Many of our core assessments, like papers, slide IDs, reflection papers, are now easily completed by tools like ChatGPT. It’s easy to feel defeated, but that doesn’t mean we give up. It means we adapt. And who better to do that than us? We are the creative ones, right?

Even though I now work in the Center for Online Learning, I started my professional career teaching Art History and I still teach it part-time. That is to say, this isn’t just theory for me. I know what it’s like to work with students and navigate the real-time challenges AI introduces. That firsthand experience has helped me reimagine how a humanities course can embrace AI as a learning partner by building strong frameworks that set students up for meaningful, authentic learning.

This post offers a practical, flexible model for integrating AI into a lower- or mid-level humanities course—using Art History as an example. The goal isn’t to overhaul your course overnight, but to help you make thoughtful adjustments that acknowledge the presence of generative AI while still supporting meaningful learning, academic integrity, and student growth. Each suggestion is designed to be immediately applicable, adaptable to your teaching style, and grounded in real classroom experience.

What an AI-Aware Shift Could Look Like

Read on for explanations on each of these AI-aware assessment approaches.

Comparison of “No AI” to “AI-Aware” Assessements

Traditional (No AI)	AI-Aware Approach
4 Knowledge Check Assignments	4 Knowledge Check Assignments with AI-Assisted Learning
3 In-Class Exams	3 Pre-tests (Unassisted Practice) plus 3 In-Class Exams
1 Major Paper	1 Semester-long, Creative Project with AI Integration

Offer Optional AI Use for Low-Stakes Formative Assignments

One simple but worthwhile shift is to make AI use optional on formative assessments, such as short quizzes and other low-stakes assignments intended to help students practice and reinforce what they’re learning. It is true that most students, when given the option, will choose to use AI in these cases if it helps them complete tasks more quickly, but ideally, they’ll also begin to notice that over-relying on AI often limits their understanding.

For example, a student might use ChatGPT to help answer a quiz question about the difference between Romanesque and Gothic architecture. While the AI response might be accurate, the student could struggle to recall the key features later during a timed image exam (more on that in a bit). That experience reveals an important insight: using AI doesn’t replace the need to actively engage with the material.

In this way, these tasks can be spaces for experimentation and honest self-assessment. By allowing AI without requiring it, you create a transparent and flexible learning environment that acknowledges modern tools while still encouraging students to take ownership of their learning.

Introduce a Low-Stakes Practice Exam

Another helpful strategy is to offer a practice exam with unlimited attempts, graded for participation rather than accuracy. This gives students a low-pressure way to test their knowledge and uncover any gaps. Just as importantly, it helps highlight who may be relying too heavily on AI. In other words, students who can’t complete the practice on their own are unlikely to succeed on the actual exam.

In my own Art History classes, for example, I’ll include slide identifications or attribution questions that mirror what they’ll see on the real exam. It’s a safe space to practice skills like visual recall and comparison, skills that don’t transfer well when outsourced to AI.

Many students misuse AI not out of laziness or bad intentions, but because they’re unsure what is and what is not allowed from one class to another. A practice exam gives them space to figure that out. It lets them test their actual understanding without penalty. They should be encouraged to attempt the practice exam without AI to get an honest read on what they know. In this way, the practice test becomes more than a study tool, like a reality check that encourages more intentional, responsible AI use.

Use In-Person or Proctored Exams for Summative Assessment

For summative assessments, consider keeping a traditional, closed-book format—whether that’s an in-person image exam, a proctored online version, or handwritten short responses. These approaches are already familiar in Art History and many other humanities disciplines. I couldn’t count the times I sat in a dark room, illuminated by the light of the slide projector, racking my brain trying to remember artist names, artwork titles, and dates.

In-person and proctored exams help ensure that students are internalizing foundational knowledge and that they can demonstrate it without digital assistance. These exams reduce the risk of cheating, reinforce the need for actual content mastery, and remind the student that not all assessments are AI-compatible. And beyond accountability, the act of recall under pressure activates the kind of learning that strengthens memory, attention, and long-term understanding.¹

---

Replace Traditional Papers with Creative, High-Stakes Deliverables

The reign of the final paper in humanities is over. May it rest in glorious peace.

Instead of relying on the ol’ end-of-semester research paper, think of how you could translate that paper into something more creative, like a video presentation, semester-long project, or even an oral exam. All of these formats still allow for AI-assisted brainstorming or organization, but they require students to actually understand the material in order to communicate it clearly. If they don’t, it becomes obvious. This taps into what I think of as the “personal-stakes” effect. Students are far less likely to cut corners when they have to speak in front of a professor or their peers, for fear of looking foolish.

I actually tried something like this in one of my Art History classes, even before the rise of ChatGPT. Several years ago, I replaced the final paper with a semester-long group project in which students curated their own themed exhibition. They selected the artworks, chose a location, and planned every detail—from object labels and layout to interpretive materials. The project was scaffolded throughout the semester, with each stage assessing different aspects of learning. It gave me the chance to provide ongoing feedback, and it gave students time to deepen their understanding week by week. The semester ended with students pitching their exhibits as if to potential investors—a format that pushed them to synthesize and present their ideas clearly. It remains one of my favorite teaching experiences.

Adjust Expectations for Upper-Level Courses

These recommendations are intended for lower- and mid-level humanities courses. In upper-level courses, assignments can become more nuanced and intentionally AI-aware.

For example, ask students to compare their own interpretations of artworks or texts with AI-generated responses. Have them revise and annotate AI outputs to evaluate accuracy, depth, or bias. Design projects that require students to analyze how AI approaches a topic, not just use it to complete the work.

At this stage, students are better equipped to evaluate sources, question assumptions, and reflect on process. These kinds of assignments help prepare them not only for final exams, but for a world where AI will continue to shape how we write, think, and communicate. They’re no longer just content consumers. Instead, they’re developing the skills to be thoughtful critics and creative contributors.

Final Thoughts: From “No AI” to “AI-Aware” Classrooms

This approach isn’t about handing over the teaching to AI. It’s about helping students learn to think critically while using it. When we design assignments that treat AI as a tool rather than a shortcut, we support academic integrity, encourage student autonomy, and align our courses with the realities of the world students are preparing to enter.

Humanities courses are still essential. But the way we teach them needs to adapt.

If you're unsure how to move forward with AI in your classroom, consider this a starting point. With a few practical changes, you can build a learning environment that is flexible, thoughtful, and ready for what comes next.

Need help getting started? Reach out to the Center for Online Learning. We’re here to support you—whether you want to brainstorm one assignment or rethink your whole approach.

Footnotes

1. Henry L. Roediger and Andrew C. Butler, “The critical role of retrieval practice in long-term retention,” Trends in Cognitive Sciences 15, no. 1 (2011): 20-27, https://www.sciencedirect.com/science/article/abs/pii/S1364661310002081?utm_source=imaginepro.ai; Jeffrey D. Karpicke and Janell R. Blunt, “Retrieval practice produces more learning than elaborative studying with concept mapping,” Science 331, no. 6018 (2011): 772-775, https://www.science.org/doi/abs/10.1126/science.1199327?utm_source=imaginepro.ai. ↩