Preserving Academic Integrity: The Role of Generative AI and its Misuse

Exploring the Misuse of AI in Academic Dishonesty

A study by Intelligent.com discovered that nearly one-third of university students have used ChatGPT for academic purposes in the past year. Interestingly, while about half of these students view the use of AI tools like ChatGPT for academic work as a form of cheating or plagiarism, 20% of students confessed to using them, as per a survey by Best Colleges.

Generative AI tools offer students multiple ways to cheat by producing unique content for unethical use. These include leveraging AI for plagiarism, enabling students to generate essays or assignment responses that seem original and evade plagiarism checks. The automatic generation of text allows for the creation of answers, posing difficulties in evaluations that require personal comprehension. The misuse of these tools underscores the need for educational institutions to take steps to address and prevent such dishonest practices.

Crafting Assessments to Thwart Generative AI

To effectively deter cheating, it’s crucial to design assessments that discourage dishonest behavior. This involves creating evaluations that require higher-order thinking skills, such as analysis and synthesis, rather than just memorization of facts. Tactics include assigning tasks like case study analysis and incorporating open-ended questions to prevent rote learning or copying. The inclusion of visual or interactive components, like images or graphs, adds a layer of complexity and prevents copying into AI tools, necessitating analytical abilities that AI cannot easily replicate. Moreover, conducting regular assessments throughout the course lessens the temptation to cheat with AI tools, as it curbs procrastination and reduces the stress associated with high-stakes tests.

The Role of edaptAI in Preventing Cheating

edaptAI’s constructed response item templates allow designers to create assessments that necessitate unique, student-generated responses, enhancing the integrity of the evaluation process and ensuring a more accurate measure of individual learning outcomes. With selected response question types, students can easily cheat by posing questions to generative AI chatbots and receiving answers. edaptAI acts as an effective deterrent to this type of cheating by designing assessment items that require higher-order thinking with constructed response item types. This method makes it difficult, if not impossible, for individuals to find answers using generative AI tools. Additionally, edaptAI’s intuitive features facilitate the easy integration of multimedia into items, strengthening defenses against generative AI and fostering a comprehensive and robust assessment experience.