In a rapidly evolving digital landscape, Artificial Intelligence (AI) has permeated almost every aspect of our daily lives, from the way we communicate to how we shop, work, and even entertain ourselves. As AI continues to advance, it is safe to assume that in the near future, AI will have influenced almost all writing tasks, from drafting emails to generating reports and creating content. Educational institutions are losing faith in AI detection software, and instead we should provide alternative approaches and guidelines for educators and students to navigate this new information age..
High Frequency of False Positives: AI detection software is highly unreliable and often produces false positives. Research found that over half of the essays written by humans were incorrectly identified as AI-generated. This lack of accuracy undermines the credibility and usefulness of such tools, as it suggests that a significant proportion of genuine, human-created content is being wrongly flagged.
Lack of Transparency: There is a significant lack of transparency in how AI detection tools operate. For example, OpenAI, which discontinued its own AI detector due to unsatisfactory results. And Turnitin has been vague about the tool's functioning, stating only that it searches for patterns typical in AI writing without specifying what those patterns are. This lack of clarity makes it difficult for users to trust or understand the results generated by the tool.
Higher Institutions of Learning Are Losing Faith and Turning Them Off: The unreliability and lack of transparency of AI detection software have led to a loss of faith in these tools by educational institutions. For example, Vanderbilt University decided to disable AI detection software, expressing doubts about its effectiveness and suitability for academic use. Many other Universities are doing the same.
Implicit Bias Against AI Writing: The use of AI detection software implies that all AI-generated writing is inherently negative or undesirable. This assumption fails to consider the potential benefits and legitimate uses of AI-generated content, such as aiding in the creation of draft materials, generating ideas, or assisting with language translation.
Discrimination Against Non-Native English Speakers: Research indicates that AI detection software can unfairly discriminate against non-native English speakers (Sample, 2023). According to this research, LLMs like ChatGPT are trained to produce text with low perplexity, meaning the text is highly probable and uses common words in familiar patterns. As a result, non-native English speakers, who are more likely to use simpler language and common words, face a higher risk of having their work misclassified as AI-generated. This raises serious concerns about the fairness and ethical implications of using AI detection software in a diverse, globalized world.
What Can Teachers Do Instead?
Have Conversations About Appropriate Use of AI: It is crucial for teachers to have open and honest conversations with their students about the appropriate use of AI as accepted by their class. This includes discussing the potential benefits and drawbacks of using AI, setting clear guidelines on when and how it is acceptable to use AI in their work, and encouraging critical thinking about the implications of AI-generated content.
Collect Writing Samples Throughout the Year: Teachers should periodically collect small writing samples from their students throughout the year. These samples can serve as a benchmark to compare against any questionable responses. If a piece of work significantly deviates from a student's usual style or quality, it may warrant a conversation with the student about their use of AI or other resources in their work.
View Collaboration with AI as Similar to Working with a Tutor: It is important to remember that collaborating with AI should be viewed similarly to working with a tutor or asking for parental help. Just as students may seek assistance from tutors or parents to understand a concept better or to improve their work, AI can serve as another tool to aid in their learning and development.
Capture Student Reflections on the AI Process: When allowing the use of AI, teachers should make sure to capture student reflections on the process. This can include asking students to document their thoughts on how using AI impacted their work, what they learned from the experience, and how they might use AI differently in the future. This reflective process can help students develop a more thoughtful and critical approach to using AI in their work.
Allow Idea Generation and Feedback: Teachers should encourage students to use AI for idea generation and then offer feedback on the AI-generated content. This approach allows students to leverage the creative potential of AI while also encouraging them to think critically about the content generated and to make necessary revisions to ensure the final work is of high quality and originality.