[AI tools] are a great starting point for many tasks but not the endpoint.
- Dr Sean van der Merwe, Mathematical Statistics and Actuarial Science
Always remember when using AI you still need to adhere to the ethical standards and policies of the UFS. Submitting work generated entirely by AI without permission is considered academic misconduct. Here are a few tips on how to use AI responsibly:
If you are a lecturer, consider including information on how your students may use AI (or not) in your course guide. For example, if you want them to cite AI tools, communicate this and provide the citation format you want them to use. If you are using AI detection tools, don't rely on % alone. If the tool detects 100% AI generated content, there is most likely a problem. And if it detects 0%, then most likely there is no problem. But AI detection tools are not always consistent or accurate (yet).
A note on AI-generated references: always check the references! Do they actually exist? Is the reference relevant? Some references are generated based on one keyword in the referenced work and is not relevant to your work.
There are tools to help you detect AI-generated text. Studies have found discrepancies between these tools that emphasise the growing challenge in AI-generated content detection and its implications for plagiarism detection. A holistic approach for academic integrity issues is a better option than depending on AI detection tools. It is better to also include manual review and contextual considerations when plagiarism is suspected.
If you are looking for AI detection tools, here are a few: