AI should be used responsibly, with integrity, and, of course, ethically. As the sphere of AI develops and education institutes tailor their responses, we must consistently assess how this rapidly changing field may affect research, learning, teaching, and assessment.
Learners at RCSI are advised to always discuss the use of AI with their lecturer and/or refer to course guidelines on this subject. Depending on the specific assignment and/or course, you may be asked not to use AI, while in others, AI use may be part of the overall assignment design. Communication is key!
You should not use AI this way because most institutes would view an entirely artificially created written assignment as plagiarism because it is not your own work.
Expect more tools to identify whether or not work is the result of an AI. Here are some examples of tools:
An option for teachers to monitor student work for plagiarism and cheating. It offers an easy-to-use interface to help teachers quickly detect any red flags in student work.
It claims a high accuracy rate and can detect AI or plagiarised content in multiple languages, including French, Spanish, German, and more.
GPTZero is trained to detect ChatGPT, GPT4, Bard, LLaMa, and other AI models.
This is an online demo of the GPT-2 output detector model based on the Transformers implementation of RoBERTa.
Proof-of-concept demonstration of using a ViT model to predict whether an artistic image was generated using AI.
This is an all-in-one tool that comes with regular plagiarism detection. It uses advanced algorithms to detect AI-generated content and produce unbiased similarity reports.
It requires a subscription, and there have been mixed reviews, but many universities use Turnitin's other features.
Due to its accuracy, educators recommend it. It uses advanced algorithms and natural language processing models to quickly identify instances of AI plagiarism, regardless of the assignment size.
Who holds the copyright if you create something using an AI such as ChatGPT? Check the Terms of Use for the AI you use for some guidance. For example, the Terms of Use for OpenAI (which includes ChatGPT) include the statement that "you may not represent that output from services was human-generated when it was not."
A few important readings on AI and copyright:
ChatGPT and Generative AI have raised concerns about academic integrity, their application, and even their definition. You can recap the RCSI academic integrity principles here.
Generative AI tools can undermine academic integrity by allowing learners (and researchers) to submit entire essays, projects, research, etc., that are not their own. This could be considered plagiarism. Furthermore, over-reliance could harm the development of key skills like critical thinking.
At present, any AI generative tool should be employed with balance in mind.
Images in a graphic novel created using the artificial intelligence system Midjourney should not have been granted copyright protection, the U.S. Copyright Office said (US Copyright Office Letter).
Brittain, B. (2023, February 23). AI-created images lose U.S. copyrights in test for new technology. Retrieved February 24, 2023, from Reuters website: https://www.reuters.com/legal/ai-created-images-lose-us-copyrights-test-new-technology-2023-02-22/
When might the use of AI tools constitute plagiarism? If you put in a prompt for an AI tool and turn in the results as your own, isn't that the same as presenting someone else's work? Isn't that plagiarism?
Lyda, M. (2022, December 15). Plagiarizing Chat GPT - Is it illegal? YouTube. (auto-generated captions) [4:43 minutes] [Chat GPT wrote the entire script]
Stern, J. (2022, December 21). Cheating with ChatGPT: Can openAI’s chatbot pass AP lit? Wall Street Journal. YouTube [6.58 minutes]
AI tools collect user information that cannot be deleted - and may share information with third-party vendors. Consider this privacy concern as you use these tools.
AI can produce sexist, racist or otherwise biassed results
AI can produce erroneous or misleading information, including making up citations for articles and books that do not exist!
AI might not include up-to-date information.
AI may produce output that is not unique. See OpenAI (which includes ChatGPT) 's Terms of Use
See OpenIA.ChatGPT: Optimizing language models for dialogue and scroll down to "limitations."
Carpenter, Todd A. (2023, January 11). Thoughts on AI's impact on scholarly communications? An interview with ChatGPT. Scholarly Kitchen.
Chen, Brian X. (2022, December 21). How to use ChatGPT and still be a good person. New York Times. Chen notes some of the limitations of ChatGPT including a reminder that the results may look plausible and yet be incorrect(!) Results may be biased based on the data, text, images given to create the AI.
Koplin, J., & Hatherley, J. (2022, December 19). Guest post: It has become possible to use cutting-edge AI language models to generate convincing high school and undergraduate essays. Here's why that matters. Practical Ethics.
Markkula Center for Applied Ethics. Internet ethics (Includes AI)
Nolan, Beatrice. (2023, January 15). This man used AI to write and illustrate a children's book in 1 weekend. He wasn't prepared for the backlash. Business Insider.
Thompson, Ben. (2022, December 5). AI Homework Stratechery. Good descriptions of ChatGPT creating some misleading and incorrect information for a homework assignment.