Resources for Detecting Plagiarism & Artificial Intelligence (AI)
A specifically designed Turnitin.com Drop Box can be placed in any Brightspace course for students to submit their papers. An Originality Report will be generated for each submission, the result of comparison between the text of the submission against the search targets including billions of pages of active and archived Internet information, a repository of works previously submitted to Turnitin, and a repository of tens of thousands of periodicals, journals, and publications. The sources for the latter include Academic One File, Emerald, ABC-Clio, Sage e-Reference, and CrossRef.
Many faculty members are currently using it as a help in editing papers. Faculty are registering their classes, setting up assignments and giving students time to submit drafts of their papers. Students then have an opportunity to redo their papers before submitting them to their professors at the final due date. This is a powerful demonstration to students that faculty do indeed care about the quality of their papers and the students' credibility.
Faculty may submit single individual papers to Turnitin without using a Brightspace interface. For further assistance with Turnitin, please call the CTE Department at (570) 941-4038.
Turnitin Enabled Dropbox Quickstart
For more information about Turnitin, go to their Web site:
AI Detection is an Indicator, Not an Indictment
Even though AI‑detection tool developers are confident in the accuracy of their solutions, it is important to recognize that their real‑world performance may not always align with those expectations. Several factors can contribute to false positive reports. Spell check support, such as the free version of Grammarly or the spelling suggestions in Word, can be flagged as AI use. Converting a document to a PDF can also trigger unseen formatting that suggests AI use. AI detectors can also flag writing from non-native speakers, neurodivergent students, and even graduate-level students who are very strong academic writers. Unfortunately, there is no simple or fully dependable approach for determining whether generative AI contributed to a piece of academic writing.
How Can AI Detection Be Used?
If you see a high AI detection score on student work, do not rely solely on that score. Consider other red flags:
- Check style and tone. A useful question to consider is whether the student’s writing is consistent throughout the paper or whether there are noticeable shifts in tone, complexity, or style. Some educators refer to this as the “AI hamburger.” The work begins in a manner typical of the student, but the middle section suddenly becomes unusually polished or stylistically distinct from the surrounding text. Such abrupt changes can raise questions about authorship, though they are not definitive evidence on their own.
- Verify citations. Encourage students to cite their claims and provide the sources that support them. If this feels too time‑consuming to review, you might consider limiting students to a curated set of sources that you already know well. Look for in-text citations that are missing elements, like page numbers. Look for citations that do not really connect to class content or are from a paywall-protected source not available to students at the U of S. Watch out for sweeping claims without textual evidence.
- Look for technical indicators. Watch for very mechanical language and limited use of personal analysis. Overuse of bullet points, icons, or em dashes can also indicate AI use. You may also notice dependence on tricolon usage or the insertion of icons.
- High-frequency AI Words. There are common words used in AI-generated content. These may include the following words or phrases.
- Delve into
- Underscore
- Pivotal
- Realm
- Harness
- Illuminate
- Revolutionize
- Cutting-edge
- Game-changing
- Scalable solution
- That being said
- At its core
Seeing these words or phrases does not necessarily indicate AI use, but may be another clue to consider with the whole picture.
- Compare to previous writing. If the structure of the course allows, consider avoiding a single high‑stakes paper as the only writing assignment. Having multiple writing samples over time gives you a clearer sense of a student’s typical voice and technique, making it easier to recognize work that differs significantly from their usual style.
- Talk to the Student. A direct conversation will help you gauge if the student understands what they were writing about. They should be able to explain their reasoning behind the work and discuss their research or writing process.
Suggested Assignment Structure Adjustments
- Require that students submit early drafts or share their document history.
- Request that students visit the Writing Center and have the staff there email the client notes to record their progress/growth.
- Consider a first draft, outline, or thesis development that is written in class and submitted before submitting the final product.
- Include a brief oral defense of the paper that students present in class or share digitally.
- Reflection logs can also be incorporated while students are working on a writing project. These are especially effective when maintained as a collaborative document that you can review and spot‑check periodically throughout the process.
These strategies may seem like additional work, but you do not necessarily need to evaluate the early drafts. Their primary purpose is to provide context should any concerns arise about the final submission. Focusing on the writing process alongside the final product can also benefit students by reinforcing stronger habits and deeper engagement with their work.
AI detection is now included if you submit an assignment through Turnitin. Even though AI‑detection tool developers are confident in the accuracy of their solutions, it is important to recognize that their real‑world performance may not always align with those expectations. There are things that you should consider to safeguard against false positive
AI detection reports.
What can trigger false positive AI detection reports?
- Spell check support, such as the free version of Grammarly or the spelling suggestions in Word, can be flagged as AI use.
- Converting a document to a PDF can also trigger unseen formatting that suggests AI use.
- AI detectors can also flag writing from non-native speakers, neurodivergent students, and even graduate-level students who are strong academic writers.
How can you establish ownership of your writing?
- Create time-stamped brainstorming notes even if they include early thoughts or messy ideas. For example, create notes on a Google document, so the creation date is saved. This would also work for a Wod document saved to your OneDrive, or using a OneNote notebook, which will also record the creation time.
- Develop an annotated research log with summaries in your own words of the content you plan on using from a specific resource. You may also wish to make notes in your lecture notes, indicating content that may support your writing.
- Save a copy of your first draft or early outlines. This shows the origins and ownership of your early work.
- Write a paper in Word, saving the document to your OneDrive so you have a document history that shows your writing progression with time stamps.
- Writing in Google Docs will also generate a document history. However, you may need to convert it to Word or a PDF to submit it. Keep the original in Google Docs, so you retain the document history for future reference.
- Keep some brief notes as you work or use the comments feature to leave yourself notes as you go. These habits can help show the process of your writing decisions.
- Visiting the Writing Center will also create a documented history of your progress and growth.
Know the class policy
- Check the course syllabus to ensure you understand the policy for AI use in the course or for specific assignments.
- If it is not clear to you, ask your instructor.
- If you can use AI, cite it properly. This includes disclosing the prompt you used, the specific tool, and the day you accessed it.
Finally, if your instructor asks to discuss your writing, try not to panic. We are all learning how to navigate the rapidly changing role of AI in education, and it’s important to give ourselves and each other some grace as we figure it out together. Stay calm, be open about your writing process, and approach the conversation as an opportunity for clarity rather than a confrontation. Bring the documentation for your process with you.
What is AI?
AI is a field of computer science that focuses on developing intelligent machines with human-like thoughts and behaviors. Artificial Intelligence (AI) has now been used in a wide range of applications, from self-driving vehicles to credit card fraud detection and medical notetaking. AI excels at repetitive, detailed-orientated tasks, and this technology has grown in its capacity to reason, self-correct, and learn. Over the past decade, AI reliance has become increasingly prevalent with the integration of AI into everyday tasks. Some common functions include text-prediction, language translation, text summation, and content generation. Other common capabilities include generating code or creating artwork.
AI functionality continuously develops with each model demonstrating more human-like behavior than the last. It has also become more immediately accessible to students with programs like ChatGPT 3.5 requiring nothing more than a user ID and password. With these advancements, professionals in higher education must consider how to adapt their pedagogy and perspectives on academic integrity so that they may continue providing students with a high-quality education.
Risks of AI Usage
AI dependency is the greatest risk posed by AI as it continuously enters new contexts within our daily lives. If one becomes overly reliant on AI by using it as a substitute for daily tasks, rather than an assistant in completing those tasks, it limits an individual’s opportunity to practice and develop certain skills. For example, if you become accustomed to using AI to answer your emails, you may not be able to respond as efficiently and effectively without the tool over time. However, there are also several risks associated with using content generated from AI:
- Responses are limited: While disciplines depend upon empirical research to expand knowledge, AI uses a body of pre-existing text to combine ideas or generate ideas, but it does not have the capacity for original thought, so it cannot create new knowledge in the same way. AI may appear to be capable of original thought, but it merely combines preexisting information or uses statistical information or rule-based systems to present new content from previous data. Moreover, AI can usually only access text back to a certain date. For example, ChatGPT can only access information prior to 2021, when it was trained, so it cannot respond to current events accurately. Likewise, Chatbots lack the capacity to truly think at the same caliber as humans, though they may sound convincingly human in their responses. For this reason, AI responds more accurately to surface-level questions with objectively correct responses. More subjective questions may result in unpredictable or less accurate responses.
- Responses must always be corroborated with additional sources: There are documented instances where AI “hallucinated” or fabricated answers that it could not generate. The danger is that the responses are presented in an elevated mannerism with extensive explanation and those unknowledgeable about the subject may be misled. Similarly, AI that has been trained on biased content, like texts from the internet, will reproduce the bias in their responses, spreading societal divides.
- Responses may plagiarize or use data unethically. Remember, the information that you enter into an AI platform may be used to train future AI, so it’s crucial that you do not input copyrighted material or your own original work. It is also not clear what data was used to train some AI platforms, and previously, some AI used copyrighted materials without the correct permissions. There is even a case where AI was trained from private medical records.
Explore AI Tools
The best way to educate the next generation of professionals who will use AI in their daily life is to educate yourself on its capacity and functionality. It is important for educators to see how AI may enhance a student’s learning experience.
Please note: The resources listed on this page are for informational purposes only and do not reflect any official 糖心APP endorsement or policy on the use of AI tools.
(Boston College Center for Teaching Excellence)
(Carleton College)
(Bowling Green State University)
(UNC, School of Global Public Health)
(Brandeis, Center for Teaching & Learning)