Detect AI-Generated Content in Student Assignments: A Teacher’s Experience
As a teacher, ensuring students engage deeply with course material and submit original work is a core goal of education. However, with the rapid advancement of artificial intelligence (AI) technologies, particularly the widespread use of generative AI tools like ChatGPT, students are increasingly incorporating AI-generated content into their assignments. This trend is especially evident in PowerPoint presentations (PPTs), which often include text, images, and design elements that AI tools can quickly generate or enhance. This poses new challenges for maintaining academic integrity while prompting me to find ways to ensure students are genuinely learning rather than relying solely on technology.
In this context, I discovered pptdetector.com, a tool designed to analyze PowerPoint files and detect whether their text content was generated by AI. By integrating this tool into my teaching practice, I was able to identify AI-generated content in student assignments and use the findings as an opportunity to engage in meaningful conversations with students, guiding them to use AI appropriately and fostering their critical thinking and originality. Below is a detailed account of my experience using pptdetector.com and how it has enhanced my teaching effectiveness and promoted student learning.
Discovering pptdetector.com
I first learned about pptdetector.com from a colleague who had used similar tools to check whether students’ written assignments were AI-generated. Inspired by this, I began exploring the tool’s potential for PowerPoint assignments. The core function of pptdetector.com is to analyze text content in PPT files and generate a report indicating the percentage of text likely produced by AI. This feature is particularly useful for detecting whether students have overly relied on AI tools like ChatGPT to create their presentations.
After understanding the tool’s principles, I decided to pilot it in one of my courses to evaluate its effectiveness in a real teaching environment. My goal was not only to detect AI-generated content but also to use the results to communicate with students, helping them understand the importance of academic integrity and guiding them to use AI as a learning aid rather than a complete substitute.
Implementation and Specific Findings
In the Spring 2025 semester, I implemented pptdetector.com in a course with 50 students. The course required students to submit a PowerPoint presentation as part of their final project, covering key course themes. To thoroughly assess the originality of their work, I uploaded all 50 PPTs to pptdetector.com and analyzed the generated reports.
The results were as follows:
AI-Generated Content Proportion
Number of PPTs
Percentage
Over 50%
10
20%
20% to 50%
15
30%
Below 20%
25
50%
These findings highlighted a significant issue: 20% of students submitted PPTs with over half of the text likely AI-generated, and another 30% had used AI tools to some extent. This suggested that a considerable portion of students may not have fully engaged with the course material, instead relying on AI to complete their assignments.
Impact on Teaching
The detection results provided valuable insights for my teaching. By reviewing the reports, I could identify which students might need additional guidance and support. For those whose PPTs were flagged as having a high proportion of AI-generated content, I arranged one-on-one meetings to discuss their project content and creation process.
During these discussions, many students openly admitted to using AI tools. For example, one student explained that they used ChatGPT to generate the introduction and conclusion sections of their presentation because they were unsure how to start or felt overwhelmed by assignment pressures. I used these opportunities to explain that original work is not only a requirement for academic integrity but also essential for developing critical thinking and communication skills. I also guided them on how to use AI as a supportive tool, such as for brainstorming or organizing ideas, rather than copying AI-generated content directly.
In one specific case, a student’s PPT was flagged as having 60% AI-generated text. Through our conversation, I learned that they had used ChatGPT to produce slide content due to time constraints. I encouraged them to revise the assignment, ensuring the content reflected their own understanding, and provided specific writing and research advice. In their resubmitted version, the PPT contained only 10% AI-influenced content, and the work was more personalized, demonstrating a deeper engagement with the course material.
Through these interventions, I not only addressed academic integrity concerns but also helped students build greater confidence and competence in their learning. Ultimately, the overall quality of assignments in the course improved, and students demonstrated deeper understanding and stronger communication skills in subsequent oral presentations.
Student Responses
Students’ reactions to the use of pptdetector.com varied. Some students understood the detection results and appreciated the opportunity to learn from their mistakes. For instance, one student, after learning their PPT was flagged, proactively requested to resubmit and noted that the experience taught them how to better organize their thoughts. However, others initially viewed AI use as a reasonable time-saving strategy, particularly under heavy academic pressure.
To address these differing perspectives, I organized a class discussion on academic integrity and AI use. I explained the importance of original work for personal growth and emphasized that AI tools should serve as aids, not substitutes. Through these discussions, most students gradually accepted the necessity of detection and expressed a willingness to prioritize originality in future assignments. Some even noted a newfound awareness of AI tools’ limitations, such as their tendency to produce content lacking personal perspective or in-depth analysis.
Government and Academic Perspectives
At the governmental level, the use of AI in education is gaining increasing attention. The Victoria Department of Education, in its guidance, explicitly recommends that schools update their “acceptable use agreements” to incorporate norms for generative AI tools (Promoting Academic Integrity). The guidance emphasizes that schools should clearly define academic integrity, plagiarism, and cheating, and provide age-appropriate citation guidelines for AI tools, such as using the APA format from the American Psychological Association. Additionally, it advises teachers to determine whether and how AI can be integrated into assignments based on assessment goals, such as prohibiting AI use in certain contexts (e.g., in-class tests) while encouraging it in others (e.g., brainstorming).
This policy framework aligns closely with my teaching practice. By using pptdetector.com, I was able to implement these guidelines, ensuring students understood the boundaries of AI use and promoting academic integrity through clear feedback and guidance.
From an academic perspective, a paper on arXiv provides critical insights into the effectiveness of AI detection tools. Perkins et al. (2024), in their study titled “GenAI Detection Tools, Adversarial Techniques, and Implications for Inclusivity in Higher Education,” found that current AI detection tools have limited accuracy (GenAI Detection Tools). Specifically, these tools achieve an accuracy of 39.5% for detecting unmodified content but drop to 17.4% when analyzing content modified to evade detection. The authors caution that, due to low accuracy and the risk of false accusations, these tools are currently unsuitable for determining academic integrity violations. They recommend a more holistic approach, such as using detection tools non-punitively to support student learning.
This research significantly influenced how I used pptdetector.com. I recognized that detection results should not be the sole basis for penalties but rather a starting point for communication and guidance. By combining detection results with student interviews and assignment resubmissions, I could assess students’ efforts more fairly while avoiding misjudgments due to the tool’s limitations.
Conclusion
My experience with pptdetector.com has deepened my understanding of AI’s role in education. While the tool is not flawless and requires cautious use, it has provided a valuable perspective, helping me identify challenges students face in their learning process. Through open discussions about detection results, I have not only upheld academic integrity but also empowered students to develop greater motivation and competence in their learning.
Moreover, my experience aligns with broader governmental and academic trends. Institutions like the Victoria Department of Education emphasize the need for clear AI use policies, while researchers urge critical consideration of detection tools’ limitations. As AI technologies continue to evolve, educators must continually adapt teaching methods to ensure students leverage AI while cultivating critical thinking, creativity, and independent research skills.
I encourage other teachers to explore tools like pptdetector.com as part of their efforts to maintain academic integrity and promote student learning. By combining these tools with clear policies and personalized guidance, we can help students become more capable and responsible learners in the AI era.