Does AI Cause Students to Plagiarize?

Since ChatGPT was opened to the public in November 2022, academia has struggled to respond. On one hand, AI represents great opportunities for educators and students. On the other hand, it represents a tempting shortcut and a new way for students to plagiarize or cheat.

Educators overwhelmingly believe that AI has increased cheating and will continue to do so. Teachers are increasingly turning to AI detection tools as students ramp up their usage of AI to aid with assignments.

But this raises a simple question: Does AI directly cause cheating? Are students who use AI systems more likely to cheat? A recent study from the University of the Basque Country examined this issue and found some interesting results.

It found that, while AI usage makes students slightly more likely to plagiarize, the effect is “virtually non-existent” compared to other factors.

However, this study doesn’t exonerate AI. Instead, it reminds us that the causes of academic cheating are complicated and that AI usage is just one factor among many.

Understanding the Study

The researchers surveyed 507 social science students at Spanish universities. Most of the students are working on degrees in education, while a minority are in related fields.

The researchers asked the students about their use of AI, whether they had committed plagiarism, their academic load, time management skills, and more. They then analyzed the data to determine which attributes correlated with plagiarism and cheating.

They found that ChatGPT usage only accounted for 3.9% of the variance in plagiarism. However, a combination of ChatGPT usage, belief in a culture of cheating and a lack of motivation raised that to 28%.

As such, the researchers concluded that other factors, most notably poor time management, lack of motivation and the belief in a culture of cheating, were much more significant factors in plagiarism.

This likely isn’t a shock to people who have studied plagiarism for any time. Study after study has indicated that these and similar factors are often the primary motivators of student plagiarism.

That said, the study does have some significant limitations. First, the study only focuses on Spanish university students in social science. It’s tough to say how the findings apply to different languages and fields.

Second, the study relies on self-reported data. Such data likely minimizes the prevalence of plagiarism, though it does make it more consistent with other self-reported studies.

Despite the limitations, the study does have some interesting insights, even if they might be of limited value at this time.

Key Takeaways

The study examines a simple question, what correlation does generative AI (ChatGPT specifically) have with students who have plagiarized?

The answer is not significantly, at least when compared to other factors. Though correlation does not equal causation, this indicates that generative AI is not directly causing students to plagiarize.

However, I don’t think many people argue that generative AI directly causes plagiarism. This is similar to arguing that a word processor directly causes an essay. While it makes writing an essay easier, it’s not why people write essays.

Instead, most argue that generative AI simply makes plagiarism easier to commit and more difficult to detect. That increases the temptation for students who are already dealing with other factors.

That said, AI does have legitimate uses for students (where permitted). These include helping brainstorm ideas, checking grammar/spelling, and finding sources. So a positive takeaway is that, just because a student uses a generative AI doesn’t mean they will use it to plagiarize.

This puts the focus back on working with students and crafting policies that explain what uses of AI are acceptable (if any) and set clear boundaries.

A student using AI may not be significantly more likely to commit plagiarism than one not, but AI still needs clear boundaries and guidelines.

Bottom Line

For instructors, it’s easy to feel everything changed in November 2022. But, while the public availability of generative AI was a big deal, it didn’t change the motivations of why students plagiarize.

To be clear, generative AI makes it easier to plagiarize and more difficult to detect plagiarism after the fact, leading to an increase in plagiarism. Anything that makes an act easier and less risky will increase its prevalence.

However, the reasons why students are tempted to plagiarize remain essentially the same. A student who is confident in their abilities, engaged in the coursework, and has good time management skills is unlikely to plagiarize, AI or not.

This means that, even in the age of AI, the classic approaches to preventing plagiarism still hold weight. Encouraging good time management, giving students resources, and crafting assignments that students want to complete can still discourage plagiarism.

As discussed in February, this is just one layer in the “Swiss cheese” approach to academic integrity. However, it may be the most important one, as it prevents plagiarism before it starts.

Though AI may have changed how students plagiarize, it hasn’t changed why. That is an opportunity that educators would be unwise to ignore.