In addition to privacy, copyright, and the risk of misinformation, there are several important challenges you should be aware of when using tools like ChatGPT:

Lack of Transparency: OpenAI shares little about how the model is trained.
Ethical Burden: The moderation filter is trained by humans who have had to go through large amounts of harmful content – often under poor working conditions.
Environmental Impact: The training process requires a lot of power, water, and storage, leading to high CO₂ emissions.
Free Labor: When you use the tool, you provide data that OpenAI uses to improve the model – without receiving anything in return.
Bias in Representation: The model is trained on data that is imbalanced in terms of gender, ethnicity, and geography.
Built-in Filters: Some versions of the model may avoid responding critically about, for example, authoritarian regimes.

These issues are important to be aware of – both for ethical and academic reasons.

Warning: Do not use ChatGPT, Copilot or Gemini to "retrieve"/find source references!

AI tools like ChatGPT, Copilot and Gemini can provide you with source references that appear genuine but are actually fabricated. Using such false references is considered serious cheating in academia.

You can use these AI tools to:

understand academic topics and concepts
get help with structure, language, and formulations

If you receive source references from an AI tool, you must always:

check if the source actually exists in OriaLinks to an external site. or academic databasesLinks to an external site.
read the original source yourself before referencing it. Even if AI provides author names, titles, and page numbers, there is no guarantee that the information is accurate. There is a greater chance that it is not correct! You are responsible for verifying everything you refer to.