Navigating Citation in a World of AI-Generated Content.
Should AI-Generated Content be Considered Plagiarism?
Welcome to the Ubik Blog!!
If you haven’t already subscribed, subscribe here!
Here’s some cool stuff you should check out :)
Students say NYC school buildings need a climate change!
A parade in Austin full of creative robots and young minds.
Check out our AIED tools here:
Navigating Citation in a World of AI-Generated Content.
Should AI-Generated Content be Considered Plagiarism?
Writing school essays used to feel daunting, facing the challenge of blank pages and tired eyes. Now, with 43 percent of college students in the US using ChatGPT, the landscape is evolving. ChatGPT, an AI chatbot tool, can assist in writing and offer a starting point for ideas and drafts. While platforms like ChatGPT are revolutionary, offering a fun and interactive way to help in writing, they also challenge academic integrity by citing information without referencing specific sources (Nerdynav, 2024).
AI tools can pose challenges in attributing credit and ensuring the accuracy of information. Citing others' work for independent research projects is a core tenet of higher education. For educators, plagiarism not only shows laziness but also discredits students work. With new leaps in generative AI, citing knowledge has become increasingly difficult.
A Conversation with ChatGPT4:
Me: Can ChatGPT cite sources reliably?
ChatGPT: As of my last update in April 2023, ChatGPT cannot reliably cite sources for several reasons:
Training Data Limitations: ChatGPT's responses are generated based on patterns and information present in its training data, which includes a vast array of text from various sources up to the last update. However, it doesn't have the capability to access or reference specific sources or databases directly.
Lack of Real-Time Access: ChatGPT doesn't have real-time internet access or the ability to pull information from external databases or websites. This means it cannot fetch or reference current or specific sources for the information it provides.
Generalized Knowledge: The model provides responses based on a general understanding of topics, synthesizing information from its training data. While it can offer information that is generally accurate, it cannot specify where each piece of information was learned or provide a formal citation as a human researcher would.
Potential for Outdated or Inaccurate Information: Due to the nature of its training, there is always a possibility that ChatGPT's knowledge may be outdated or not entirely accurate. This is particularly true for subjects that evolve rapidly.
For academic or research purposes, it's recommended to use ChatGPT as a starting point or for idea generation rather than a source of citable facts. Any information obtained from ChatGPT should be independently verified and properly cited using primary sources.
What is Academic Writing?
Academic writing is used in higher education settings like colleges and universities to present research projects, theories, and arguments. Academic writing uses formal language and defined structure to communicate original ideas with evidence and analysis. To correctly compose an academic essay, writers must use citations and references to make their work credible.
Fears of Researchers
With new developments in Natural Language Processing technologies (NLP), researchers fear students and professionals will avoid the critical checkmarks that amount to original work.
Originality and creativity are central to growing as a student and a thinker outside the classroom. With platforms like ChatGPT, students can automatically generate large amounts of text based on simple prompts.
The fear of cutting corners does not mean researchers should avoid all NLP systems or Chatbot platforms. Instead, "any section of a manuscript written by an NLP system should be checked by a domain expert for accuracy, bias, relevance, and reasoning" (Kabnick et al., 2023). It is necessary to verify information from chatbot-generated text to ensure the validity of the research.
Academic writing pushes students and researchers to contribute new ideas to specific fields of study. New ideas need to be original. Mastering this writing style takes much practice over many different projects. Still, with tools like ChatGPT, students may miss out on these essential skills (Frye et al., 2023).
Why Could Using Generative AI Be Seen As Plagiarism?
Using an AI text generator for academic writing might be perceived as plagiarism within the research community because it entails presenting content, whether generated by an AI system or not, as one's own without proper citation or authorization.
In academic settings, plagiarism is a severe offense as it undermines scholarly works' credibility and intellectual integrity. Researchers employing AI text generators risk submitting assignments lacking originality, thorough research, or robust support.
Such actions can breach the ethical standards of academic writing, leading to harsh consequences like failing grades, loss of credibility, and expulsion from educational institutions (Frye et al., 2023).
Many Academic Institutions have restricted or even prohibited generative chatbots, and with good reason. There is a need for AI tools for academics or professionals to generate trustworthy text with verifiable citations. With major worries in Higher Education, teachers believe that AI tools "could proliferate plagiarism" (Nguyen, Q. H. 2023).
How Should Generative AI Be Used in Academic Writing?
Generative AI tools like ChatGPT and other NLP technologies hold immense potential to revolutionize research and foster innovation across various fields. However, their impact extends beyond just advancing academic work; these tools can spark new ideas and enhance learning experiences. While concerns about cheating and plagiarism are valid, the focus should be on responsible usage and education about these tools rather than limiting their potential due to these risks.
For students, these AI tools are more than just problem-solvers; they're a gateway to creative thinking and a better understanding of complex material. Yet, it's crucial that students learn not to accept AI-generated responses uncritically but rather use them as a starting point for deeper exploration and learning.
The ethical integration of generative AI into academia demands a concerted effort from educational institutions. By embracing these tools and incorporating them into the learning environment, we can create new avenues for creativity and innovative teaching methods. This approach echoes the early 2000s when computer education opened new horizons for students. Now, it's time to teach the next generation how generative AI can be a powerful ally in their educational journey, encouraging them to use these tools to enhance, not bypass, their learning experience.
Collaboration, NOT Automatic Generation
Designing practical, collaborative AI tools for education ensures students understand the subject matter rather than finish the assignment with the correct answers. Educational institutions are beginning to roll back their initial bans on the technology after realizing that AI tech can foster collaboration and create positive educational opportunities (Rosenblatt, 2023).
Fiig.ai by Ubik
Fiig.ai from Ubik Studio is an academic tool that helps humans collaborate with AI to understand research without sacrificing learning.
Join us
After making an account at fiig.ai, users can:
SEARCH peer-reviewed journals
Get help using AI studies findings to PLAN projects
COLLABORATE with AI to finalize reports (with citations)
Work Cited
Frye, Brian L., Should Using an AI Text Generator to Produce Academic Writing Be Plagiarism? (December 3, 2022). 33 Fordham Intellectual Property, Media & Entertainment Law Journal 947 (2023), Available at SSRN: https://ssrn.com/abstract=4292283
Gregory E. Kaebnick et al., (2023) Editors' Statement on the Responsible Use of Generative AI Technologies in Scholarly Journal Publishing. AJOB Neuroscience 14:4, pages 337-340.
Nerdynav. (2024, January 5). ChatGPT cheating statistics & impact on education (2024). Nerdynav. https://nerdynav.com/chatgpt-cheating-statistics/#:~:text=43%25%20of%20college%20students%20have%20used%20ChatGPT%20or%20similar%20AI%20tools
Nguyen, Q. H. (2023). AI and Plagiarism: Opinion from Teachers, Administrators and Policymakers. Proceedings of the AsiaCALL International Conference, 4, 75–85. https://doi.org/10.54855/paic.2346
Rosenblatt, K. (2023, May 18). New York City public schools remove ChatGPT ban. NBC News. https://www.nbcnews.com/tech/chatgpt-ban-dropped-new-york-city-public-schools-rcna85089
Always interesting to me how AI checkers don't work very well -- i wonder if false plagiarism accusations will go up without a better framework for using AI in school/research?