Always verify the content: Think of your preferred Generiative AI as your new personal assistant (PA) who is forever in training. Your PA needs very specific instructions, and you'll to need check their work or information.
Generiative AI sometimes makes things up: That's because it's designed to write in a way that sounds like human writing. It's not designed to know facts.
1. Check for Bias and Accuracy: AI might produce biased or incorrect content. Always double-check information and sources.
Therefore, do not use AI content as a source, do not list Generative AI as a citation or use as an in-text source. Instead, use the sources you have double-checked to ensure your facts are correct. AI is not in itself a reliable source.
2. Your Judgement Matters: consider AI-generated content as a starting point, not a final solution.
Always adhere to our school's guidelines which is to acknowledge how AI was used to support your inquiries. See the page on Acknowledging Generative AI
3. Know the Limits: What are the limits of AI's content?
The following text from Copilot responds this question: "As an AI language model, I don’t have direct access to external information or databases. My responses are generated based on patterns in the text data that I was trained on. The training data includes a diverse range of sources up until 2021, but I don’t have real-time access to the internet or any specific year beyond that." (Microsoft Copilot.(2024) [Large language model]. https://www.bing.com/chat)
4. Ethical Concerns: Generative AIs training data may include Copyright Works & exploitation of AI workers
There is concern and outrage in a number of countries regarding the lack of transparency and the inclusion of copyright material (without the authors permission) in the training data used. Further, there is evidence that the creation of the training data has involved the exploitation of workers.
Students should be aware of their moral and ethical responsibilities to observe Copyright Law and to avoid breaching copyright. For this reason, it is recommended that students refer to the source rather than Generative AI text.
It is reported that OpenAI underpaid workers in Kenya, India and Uganda to label offensive data; this challenges the represenation fo this product's technological supremacy. See Perrigo, B. (2023). OpenAI Used Kenyan Workers on Less than $2 Per Hour: TIME. https://time.com/6247678/openai-chatgpt-kenya-workers/
5. Privacy Concerns
Content you enter into the chat box may be used to improve model performance, (unless you opt out https://help.openai.com/en/articles/7730893-data-controls-faq), therefore consider the sensitivity of information or images before submitting.
6. Environmental Concerns
AI servers are stored in data centres which consume vast amounts of electricity, creating greenhouse gas emissions. Large amounts of water are required for construction and to cool the electrical components. According to the International Energy Agency (IEA), a single ChatGPT request requires ten times more electricity than a Google search.
For more information on this issue: United Nations. (2025). Artificial intelligence: How much energy does AI use? - United Nations Western Europe. United Nations Regional Information Centre for Western Europe. Retrieved April 23, 2025 from https://unric.org/en/artificial-intelligence-how-much-energy-does-ai-use/
