Blue Books and Face-to-Face: When Tradition Becomes Innovation

In classrooms across America — from bustling lecture halls to quiet seminar rooms — educators are wrestling with a new kind of challenge: how to foster learning in an age where artificial intelligence can conjure essays, solve problems, and generate answers with remarkable ease. Out of this struggle, a curious revival is taking shape — one that harkens back to testing methods as old as the academy itself. The chalk-smudged desks and reflective eye contact of oral examinations are returning, not as relics of a bygone era, but as deliberate tools to guard the integrity of student assessment in the AI era.

Across campuses from the University of Wyoming to Vanderbilt and the University of California, San Diego, professors have begun administering oral exams — conversational, in-person assessments that require students to articulate their understanding face-to-face and in real time. In an educational moment marked by the rise of tools like ChatGPT, these spoken exchanges are prized for their resistance to generative AI assistance: there is simply no digital shortcut to real-time questioning and thoughtful response.

For many educators, this is more than a tactical response to cheating; it is a rediscovery of an ancient pedagogical practice. Oral examinations were once ubiquitous in places like Oxford and Cambridge, and they remain a mainstay in some countries even today. In the U.S., they had largely faded from undergraduate classrooms — reserved mostly for doctoral work or specialized fields — until the recent surge in AI-enabled submissions exposed limitations in traditional testing.

The shift is dovetailing with another old-school approach: handwritten, in-class exams completed without the aid of laptops or internet access. Sometimes administered with classic “blue books,” these analog tests remove the opportunity to paste text crafted by an AI model into a digital submission. Across institutions, instructors are increasingly insisting that students set aside digital devices and produce work on paper under supervision, making it far harder for generative tools to influence responses.

These methods reflect broader frustrations with AI detection software, which has struggled to keep pace with ever-advancing language models. For many faculty, the more reliable way to know whether a student truly grasps material is to observe their thought process as it unfolds — whether spoken aloud or produced in realtime by hand — rather than relying on post-hoc analysis of submitted text.

Students themselves report a range of reactions. Some find oral and handwritten exams more engaging, appreciating the opportunity to discuss ideas directly with instructors. Others confess to initial anxiety about the return to traditional formats, recalling the dread that used to accompany blue book essay sessions. Yet even those who once found these formats intimidating acknowledge that they can deepen understanding and lessen dependence on AI shortcuts.

This resurgence of analog assessment does not signal a wholesale rejection of technology in education. Many professors continue to integrate AI as a learning tool, encouraging ethical use for brainstorming or study support while reinforcing that uncredited AI-generated work contravenes academic integrity policies.

Instead, the revival of “old-school” exams underscores a deeper conversation about the purpose of evaluation in an AI-saturated world. When machines can produce polished essays at the click of a button, educators ask, how do we ensure that grades reflect genuine understanding rather than clever prompting? Oral and handwritten assessments, in their simplicity and immediacy, offer a clear — if demanding — answer.

AI Image Disclaimer “Visuals are created with AI tools and are not real photographs.”

Sources

• Washington Post — Professors are turning to this old-school method to stop AI use on exams

• Cornell University news — Faculty innovate with, and avoid, AI in the classroom

• Times of India — US professors bring back handwritten tests to combat AI cheating