As Artificial Intelligence continues to become more convenient and accurate to use in day-to-day life, more issues relating to AI also become more relevant. One particular instance of this is students plagiarizing AI-generated material and presenting it as their own work. T

‍

BBC reports that Cardiff University students have used Artificial Intelligence to write their essays. Students can easily input essay prompts into chatbots such as Chat GPT and quickly receive entire essays without having done any work. In an interview with Tom a student at Cardiff University, he told BBC that “for the essay he wrote with the help of AI, Tom received a first - the highest mark he has ever had at university”. He emphasized how helpful it was and how easy it was to use. It gave his information much faster he paraphrased the information from GPT, however the same cannot be said for other students. It is likely that some students copy the chatbot word for word without putting any effort in masking of obfuscating the use of it. Another student from the same university stated, "I see no way that anyone could distinguish between work completely my own and work which was aided by AI". However, there are methods that can be used to detect it.

‍

‍

Websites like Grammarly can test for plagiarism, which means if someone passes off a piece of writing as their own which copies or is closely similar to another piece of writing. This type of plagiarism is more easily detected as applications can search through billions of data on websites to check if a piece of writing matches an online one or not.

‍

‍

However, AI is different because its responses are generated in real-time and the data and information stored are not online or in any sort of database, which is why the technique for detecting AI plagiarised work is very different. In fact, the way AI is detected is through AI itself, specifically machine learning and natural language processing. According to Brendan Aw, there are two main concepts relating to these AI tools, “Linguistic Analysis: Examining sentence structure to look for semantic meaning or repetition” and “Comparative Analysis: Comparison with the training dataset, which looks for similarities with previously identified instances”. Machine learning is used in AI detection to classify and differentiate between human and AI text using features like grammar and style.

‍

The AI classifier is tasked with finding the line between AI and human text by using AI models. Some of these models include logistic regression and K-Nearest Neighbor which each have slight differences. The comparative analysis part of AI detection uses a huge dataset of AI-generated and human-generated text and uses those AI models to find patterns and differences between the two. This type of classification has many examples, for example classifying dogs from cats or tulips from roses. The method is a little different from text since computers cannot process words the same way as pixels. It requires vectorization where the computer converts the text into numbers, but the strategies are similar. 

‍

‍

Overall, the situation with AI and writing is complicated, but it is certain that plagiarism will always be prevalent. Even tools like AI detection are not completely foolproof yet which is why the best method is to prevent AI use in the first place. No matter what assignment you may struggle with, it is important not to plagiarise online documents or copy AI work. Not only is it a violation of integrity, but it is also harmful to the plagiariser’s educational experience. However, AI is a very helpful tool that should be used to help. Whether it means asking it the next step in your brainstorming or asking it a quick fact, AI chatbots can be very time-saving and useful. With new technology, there will always be people who abuse it, but understanding the line between using this technology can help all of us creatively.

‍

‍

‍

Sources

https://www.bbc.com/news/uk-wales-65167321

https://brendanaw.com/how-does-ai-detection-work

‍