AIxAI

Impact of Generative Artificial Intelligence on Academic Integrity

Concerns about AI and academic integrity are not surprising: Technology has been blamed for decades for academic integrity concerns. Sarah Eaton (2023) reminds us that calculators in schools caused widespread panic around academic dishonesty in the 1980s, and the internet had a similar impact in the 1990s. Then, in 2003, Dehn argued that “technology and other cultural factors may be synergizing to produce the perfect storm of increasing academic dishonesty” (p. 192). Later still, in 2014, Bachore claimed that “The advancement of technologies, such as cell phones, iPods, internets, has broadened the ways by which people can achieve the goal of cheating” (p. 1060). In addition, several technologies that emerged in the 2000s and 2010s, such as the grammar checker in Microsoft Word, Essay Jack (now Wize Writer), and Grammarly, have caused writing instructors great concern, not only about academic integrity, but also about students’ abilities to learn to write. So generative artificial intelligence is not the first technological tool to cause educators to question the best way to foster academic integrity in the classroom.

What Can You Do about AI and Its Impact on Academic Integrity?

The suggestions outlined in this section are not definitive, nor are they intended to be prescriptive. They are recommendations based on literature and practice regarding academic integrity and teaching and learning more generally. At the very least, hopefully they will be the beginning of a larger conversation about academic integrity and AI.

First, banning specific technologies is not an effective academic integrity strategy—and even if students were inclined to respect the ban, there are thousands of other options for people who don’t wish to do their own work: sites such as Chegg, Course Hero, Studocu, Unemployed Professors, and a myriad of essay mills and other study sites are just a few of many, many options for students to acquire and misuse unauthorized materials. Second, even if there weren’t intellectual property, privacy, and unethical data stewardship issues with AI detectors, studies show that they are also very unreliable.

However, there are ways to encourage students to embrace academic integrity in their course work. In his book Cheating Lessons, James Lang (2013) studied classroom conditions under which anyone would cheat (even you! or me!) in an effort to help faculty consider the kinds of classroom environments they co-create with their students. Here are what Lang identified as the five most impactful conditions:

If instructors foster student intrinsic motivation, learners are less likely to cheat. How do we do this? According to Pink (2009), motivation is strongest when people have autonomy, a sense of purpose, and the ability to achieve competence in skills and knowledge. In the classroom, this might mean that we provide students with autonomy, that we ensure the course content is relevant to their lives, and that we provide them with ample practice opportunities to learn more deeply.

If we create courses that allow students to achieve competence in skills and knowledge through lots of formative assessment, including practice opportunities, we prevent them from focusing only on a one-time performance. If too much of their grade rides on a single assessment or performance, students are more likely to stray from the path of academic integrity.

It will likely not surprise you to hear that it’s more likely students will cheat on a high-stakes exam than several low-stakes quizzes. Multiple low-stakes assessments foster student learning of skills and knowledge and help prepare them for higher-stakes assessment. If your course has only high-stakes assessment, students are more likely to cheat.

If students feel prepared to demonstrate skills and knowledge, they are less likely to rely on practices that stray from academic integrity to get through. Again, using formative assessment to help prepare students for summative assessment is a great way to build student self-efficacy.

Classroom culture has an impact. If you spend lots of time co-creating a positive classroom culture with your students by seeking their feedback and input when appropriate, normalizing academic integrity by having frequent honest discussions about it, students will better understand its importance and the impacts of academic dishonesty on the value of their education.


Strategies for Fostering Academic Integrity in Today’s Technology-Enhanced Classroom

What do Lang’s five principles look like in practice? In addition to the great ideas shared on the classroom ideas section of this site, they could look like the following:

Simply banning certain technologies will remove student autonomy and could lead to dishonest use of those technologies. If students themselves determine rules around AI use, they are more likely to adhere to those guidelines.

For example, where do the AI tools get their information? Are they accurately citing their sources? Who might be harmed by this lack of attribution? How might they feel if their own hard work were borrowed without attribution?

Assigning work in stages can not only help students focus on the process of learning, but it can also demonstrate to you whether they are on the right path. Scaffolded assignments can also reduce learner stress, as they reduce the grade weighting on any single piece of work, and they can help students build deeper skills and knowledge. They also prevent procrastination for a heavily weighted assessment, as several of the pieces will have been completed before the final assessment is due.

In the discussion about Lang’s findings above, additional formative assessment and fewer high-stakes summative assessments can help foster a culture of integrity in almost every instance.

Not only does this strategy encourage learners to do their own work, but it also reduces assignment procrastination and stress, which might prevent them from turning to AI to complete their work for them.

his can reduce stress for you and for students, which helps create an environment more conducive to academic integrity. Consider these questions:

  1. Where can students have choice?
  2. Can students choose due dates?
  3. Can students choose from a range of modes or formats or assessment types?
  4. Can you build in a blanket extension policy?
  5. Could you offer choice around feedback type/amount/mode of delivery?

Showing students that you value their learning rather than their final product fosters an environment where it is safe to make mistakes, resulting in less motivation for learners to stray from the path of academic integrity.

One excellent teaching and learning tool that leads to deep student learning is reflection. Research shows that reflecting on learning experiences leads to more learning than simply experience alone (Brown et al, 2014; Fink, 2013; Doyle, 2011; Ambrose et al, 2010; and others). Why is reflection so powerful? It involves several cognitive activities:

  1. Retrieval: recalling recently learning knowledge to mind (Brown et al, 2014, p. 27, 66);
  2. Elaboration: connecting new knowledge to what is already known (Brown et al, 2014, p. 89, 207; Doyle, 2011, p. 145); and
  3. Generation: rephrasing key ideas in one’s own words, visualizing what one might do differently next time; the production of new
    knowledge. Generation makes the mind more receptive to new learning (Brown et al, 2014, p. 89, 208).

    These activities are essential to help students assimilate new knowledge and make connections between existing knowledge and new learning (Bain, 2004; Doyle, 2011).

As well, it’s worth considering the place of AI in the original list of tools and technologies mentioned at the top of this page. What do they all have in common? They offer educators an opportunity to rethink the purpose and goals of their assessments. They also invite consideration about how students are best able to demonstrate their learning with integrity. For example, when tools such as Grammarly were first introduced, many educators were able to shift the focus of the goals of their assignments from helping students apply basic grammar rules to effectively demonstrating those rules in context: they used the tools to help students develop an understanding the value of editing, as grammar checkers were (and still are) not always accurate. Similarly, information currently produced by generative AI isn’t always accurate, so instructors could take the opportunity to help learners explore the value of fact-checking and editing.

AI also may allow student writers space to think more critically about argument, voice, and structure. In a recent experiment, one Twitter user, @academicswrite, shared that they invited students to use ChatGPT for a written assignment, provided that students also handed in a reflection about its use. About half the class took the opportunity to use the AI tool, but many were disappointed with the results. Some learners shared that they felt that ChatGPT had stolen their voice from them. This is a powerful revelation for emerging writers—a writer’s own voice is important.

Final Thoughts

Academic integrity concerns due to AI are similar to those that have existed for decades due to other technological innovations. AI, though, offers us an opportunity to engage in new conversations about academic integrity with students and colleagues in a meaningful way. While there is no simple solution to ensure that all students will always complete assignments with a full adherence to academic integrity guidelines, the above strategies can help.

References

Academics Write [@academicswrite]. (2023, August 8). I did a thing this term where I told my students they were allowed to use ChatGPT. The catch was [Tweet]. Twitter. https://twitter.com/academicswrite/status/1689061617495932933

Ambrose, S. A., Bridges, M. W., DiPietro, M., Lovett, M. C., Norman, M. K., & Mayer, R. E. (2010). How learning works: Seven research-based principles for smart teaching. Jossey-Bass.

Bachore, M. M. (2014). Academic dishonesty/ corruption in the period of technology: Its implication for quality of education. American Journal of Educational Research, 2(11), pp. 1060-1064.

Bain, K. (2011). What the best college teachers do. Harvard University Press.

Brown, P. C., Roediger, H. L., & McDaniel, M. A. (2014). Make it stick: The science of successful learning. Belknap Press.

Dehn, R. W. (2003). Is technology contributing to academic dishonesty? Technology and Education, 14(3), pp. 190-192.

Doyle, T., & Zakrajsek, T. (2011). Learner-Centered Teaching: Putting the Research on Learning into Practice. Stylus Publishing.

Eaton, S. (2023, March 4). Artificial intelligence and academic integrity, post-plagiarism. University World News: The Global Window on Higher Education. https://www.universityworldnews.com/post.php?story=20230228133041549

Fink, L. D. (2013). Creating significant learning experiences: An integrated approach to designing college courses. John Wiley & Sons.

Lang, J.M. (2013). Cheating lessons: Learning from academic dishonesty. Harvard UP.

Pink, D. (2009). Drive: The surprising truth about what motivates us. Riverhead Books.