Academic Integrity: Artificial intelligence (AI) tools and academic integrity

As a Marianopolis student, you are only permitted to use AI tools for course evaluations when it is explicitly stated by your teacher and with the relevant acknowledgements. 

Be sure to always check your teachers' instructions to know what is allowed or expected. There may be different expectations in every course and it is your responsibility to be aware of each of your courses' rules. If you are not sure what is allowed, you must ask your teacher.

AI text generated content is not your own work, thus submitting AI text generated content without proper citation is a violation of Academic Integrity. Please note the following section of the Institutional Policy on the Evaluation of Student Achievement (IPESA):

4.1. Definition

Academic integrity means:

4.1.1. Students are expected to submit work that is entirely their own. Any reference (direct quotes or otherwise) to another person's ideas, content, answers, or manner of expression must be cited in conformity with guidelines provided by the teacher.

4.2. General guidelines

4.2.1. Students must follow instructions and guidelines provided by their teacher with respect to the completion of all forms of evaluation.

4.2.4. Where outside sources are permitted, all sources must be identified by the appropriate means of citation as determined by the teacher (MLA citation style, APA citation style, etc.).

For more information about Academic Integrity, please refer to the Student Guide.

What is artificial intelligence?

Artificial intelligence, a discipline within computer science, involves training computers to accomplish tasks that typically require human intelligence. For example, with AI, computers can solve complex problems, communicate through natural language, and perform sophisticated visual recognition tasks. AI systems are trained on large datasets to learn patterns and relationships within the data. As it completes tasks, AI adapts and improves its functions using a process called machine learning. By continuously learning from new data and experiences, AI becomes more skilled over time, making its results more accurate and efficient.

What is generative AI?

Generative AI is an artificial intelligence system that has the capability to generate new content, such as images, text, audio, or even videos. The results of generative AI are not aimed to be accurate, but rather to mimic what humans would create. By using a mathematical system called neural networks, generative AI scans massive amounts of data in order to identify patterns and "learn" from the characteristics it finds in the data. It then is able to reproduce these patterns in its output to create new content that is human-like.

What is ChatGPT?

ChatGPT is a generative artificial intelligence chatbot developed by OpenAI. ChatGPT is based on a large language model called GPT (Generative Pre-trained Transformer). The model is trained on a vast dataset that includes a wide range of sources, such as websites, books, articles, and more, from the publicly available internet. When given a user's prompt, ChatGPT uses the patterns it has identified in its training data to predict the most likely continuation of the text given the input. It's important to note that ChatGPT is not thinking, reasoning, or researching its responses. It is simply emulating the patterns it has identified in its training dataset to provide responses that aim to be contextually appropriate and human-like, but without true understanding or comprehension.

Accuracy

Generative AI tools simply predict the most likely response to a given prompt, not necessarily the correct response. Its training dataset contains a vast amount of human-created work, which may contain inaccuracies and misinformation, which the AI tool will then reflect in its response. As a result, AI models like ChatGPT may generate responses that sound plausible but are factually incorrect, called "hallucinations". For example, when asked to provide sources on a certain topic, ChatGPT might invent sources that do not exist. The responses provided by AI models like ChatGPT cannot be trusted as truth and must be verified by reliable external sources.

Currency

Generative AI tools are trained on a fixed dataset and do not have access to real-time information or the ability to browse the internet.This means that the responses they generate are not necessarily based on the most current information available and can be outdated. For example, the current free version of ChatGPT is trained only on data from 2021 and before. Therefore, any events, developments, or information from after then is beyond the chatbot's knowledge.

Bias

As generative AI tools are trained on a set of human-created work, they may reflect the biases that are represented in that work. It is important to remember that the responses generated by AI tools are not perfect reflections of objective truth but rather a reflection of the patterns present in their training data. Biases in the training data can result in biased responses from AI models like ChatGPT, which may perpetuate stereotypes or favour certain demographics. Users should approach AI-generated content with caution and critically evaluate responses for potential biases. 

Privacy

For ChatGPT, users have to create an account in order to use the service. This means that OpenAI, the company behind ChatGPT, collects data relating to users' account data, user content, communication details, and even social media information. In OpenAI's privacy policy, they not only have the right to store this data, but also to sell it to third-party vendors. Privacy may be a huge concern for AI tool users and should be kept in mind when using its services. Furthermore, in OpenAI's terms of use, it is stated that users "must be 18 years or older and able to form a binding contract with OpenAI to use the Services". The terms also state that you must not "represent that output from the Services was human-generated when it is not".

As stated above, as a Marianopolis student, you are only permitted to use AI tools for course evaluations when it is explicitly stated by your teacher and with the relevant acknowledgements.

If you use an AI tool to perform any segment of a course evaluation without express permission and without citations, you are submitting work that is not entirely your own, which is a violation of academic integrity (see section above).

However, AI tools can be helpful as a learning tool when they are used to help you better understand course material and not to perform tasks for you. Here are some examples of ways you can use ChatGPT:

1. Brainstorming

Often, beginning an assignment or project can be the hardest part. If you are stuck for ideas, ChatGPT can create prompts to get you started. Remember that this is just a jumping-off point and not the end result. 

2. Creating a study guide

ChatGPT is trained by going through huge amounts of data and finding patterns. This means it will be helpful at synthesizing your study notes and creating a study guide for your upcoming exam. Remember to only follow it as a guide and not as a way to generate new content. 

3. Asking questions about concepts you don't understand

Sometimes, there can be difficulty in understanding concepts as they are taught in class. ChatGPT is can re-explain the same concepts in different terms, which might create a better way for you to understand the concept. Remember that the tool has its limitations and may not always be accurate, so you need to make sure that the information provided is in fact correct.

Please note that you are not permitted to share your teacher's content with any outside parties, including AI tools. Any text you submit to AI tools must be your own and not your teacher's. Please refer to the Copyrighted Material section of your course outline for more information.

Concordia University. (n.d.). ChatGPT & Generative AI. Retrieved from https://library.concordia.ca/apps/things/thing.html?thingID=22032 

Dobrin, S. I. (2023). Talking about Generative AI [Version 1.0]. Retrieved from https://files.broadviewpress.com/sites/uploads/sites/173/2023/05/Talking-about-Generative-AI-Sidney-I.-Dobrin-Version-1.0.pdf

OpenAI. (n.d.). Privacy policy. Retrieved from https://openai.com/policies/privacy-policy

OpenAI. (n.d.). Terms of use. Retrieved from https://openai.com/policies/terms-of-use

UNESCO. (2023.). ChatGPT and Artificial Intelligence in higher education. Retrieved from https://www.iesalc.unesco.org/en/2023/04/14/chatgpt-and-artificial-intelligence-in-higher-education-quick-start-guide-and-interactive-seminar/