As students increasingly rely on ChatGPT to complete their homework, take-home exams, and craft essays, educators are turning to traditional methods such as blue books-affordable, stapled examination booklets with a distinctive blue cover and lined pages-to safeguard academic integrity. These physical test booklets, first introduced in the 1920s, are experiencing a resurgence as a countermeasure against AI-assisted cheating.
The Wall Street Journal reported earlier this month that demand for blue books has surged, with campus bookstores selling them at a rate of approximately 23 cents each. This uptick in sales reflects a broader trend across universities, where faculty members seek tangible, AI-resistant assessment tools to ensure fair evaluation.
Rising Popularity of Blue Books Amid AI Challenges
Data from the past two years indicates significant increases in blue book sales: Texas A&M University saw a rise of over 30%, the University of Florida experienced nearly a 50% jump, and the University of California, Berkeley, reported an astonishing 80% growth, according to the Journal. These figures suggest that institutions are actively adopting traditional testing methods as a response to the growing prevalence of AI tools like ChatGPT.
Manufacturers and Institutional Strategies
Roaring Spring Paper Products, a family-owned company that supplies most blue books nationwide, noted that their sales have accelerated in recent years, attributing this trend to the widespread use of AI in academic settings. The company explained that blue books serve as an effective deterrent because students cannot simply input prompts into ChatGPT and submit automated responses; instead, they are required to produce handwritten work under direct supervision, making AI-assisted cheating more difficult.
Educational Responses to AI-Generated Content
In response to AI’s encroachment on academic honesty, some professors have adopted innovative assessment methods. Kevin Elliott, a lecturer at Yale University specializing in ethics, politics, and economics, recounted switching from take-home essays to blue book exams during the spring semester after noticing students’ increasing reliance on AI. He discovered that many submitted papers contained fabricated quotations from well-known philosophers-an unmistakable sign of AI involvement.
To combat this, Elliott introduced a new testing format requiring students to write essays in blue books during final exams. The approach proved highly effective, leading him to plan continued use of blue books throughout the upcoming academic year. This shift underscores a broader concern among educators about maintaining academic integrity in the age of AI.
Perceptions of AI-Driven Dishonesty
Surveys reveal that AI tools have contributed to a perceived rise in academic dishonesty. A January study conducted by the American Association of Colleges and Universities in collaboration with Elon University found that 59% of university administrators believe cheating has increased since AI tools became widely accessible. Furthermore, over half of these leaders feel their institutions lack the means to distinguish between AI-generated work and student-authored submissions.
Similarly, a survey from Take into account.com in January 2023, which included responses from over 100 educators and 1,000 students, indicated that nearly 90% of students had used ChatGPT to complete homework assignments. Specifically, 53% employed it to write essays, while 48% used it for quizzes or take-home tests. More than 70% of professors expressed concern about the potential for AI to facilitate cheating on assignments.
Balancing Innovation and Academic Integrity
Despite the challenges, some educators who restrict ChatGPT use during exams acknowledge the potential benefits of AI literacy. Arthur Spirling, a political science professor at Princeton University, remarked that while he administers supervised blue book exams, it seems counterproductive to prohibit students from using AI tools they are likely to encounter in their professional lives.
He emphasized that denying students access to AI during their education could hinder their ability to develop essential skills, stating, “It’s peculiar to prevent students from engaging with a technology that will be integral to their careers.” As of April, ChatGPT’s global weekly user base had grown to 500 million, up from 400 million in February, highlighting the rapid adoption of AI tools worldwide.
