Artificial intelligence offers many exciting opportunities in studies, research, and other work. However, there are many pitfalls you might fall into if you use the technology uncritically. You must reflect on what is good and bad or acceptable and unacceptable use. In other words, you need to make some ethical considerations.
Important questions you should reflect on
Can I input personal information into an AI tool?
Data protection requires that you have good control over the personally identifiable data you process in research or studies at University of Oslo (UiO). You cannot input personal information unless it is permitted by law, UiO has clear guidelines on this.
Do I know who is behind the AI tool, and what interests are at play?
Are the owners, or the sender, clear about how the AI tool works, how it has been trained, and how it uses the data that is inputted? If the AI tool is transparent in how it was built and processes data, it's easier to assess whether it can be used in studies and research.
Can I input someone else's text or results into the AI tool?
Other people's work may be copyrighted, and you cannot use this work as you please. If the AI tool uses the material inputted for further training of the AI tool, you cannot use the material in the AI tool.
Can I use AI tools to write my academic text and claim the output from the it as my own work?
If your subject allows it, you can use text-generating AI tools to develop ideas conceived by you, but the results should not replace your own work. If you have used a text-generating AI tool, be open about how you have used the tool. This might include mentioning the AI tool and how you used it, or providing a link to the conversation with the AI tool (chat). See a recommendation from APA on search and write.
Have you accounted for rules and guidelines on cheating and dishonesty?
Academic ethical principles require that work that is published, submitted, or to be assessed, is a result of your own intellectual effort. Artificial intelligence can assist you in formulating and drafting text, and there will be a limit to how much help you've received before you can no longer claim it as your own text. If you present intellectual work done by others, be it another author or a tool, as your own, you can be accused of plagiarism or cheating.
When writing academic text, can you be held accountable for the text you have written and the sources it is based on? Can others replicate the results you've produced using artificial intelligence?
Does UiO have rules and guidelines about cheating and dishonesty that I need to be aware of?
UiO's cheating rules state that "When you write academic texts, you should display which thoughts and reflections are your own and which you have borrowed from others' work. This allows the reader to look up the sources you've used, verify facts, and replicate your results." Fabrication of data/research data is academically dishonest.
Can I trust the answers I get from artificial intelligence-based AI tools?
Text-generating tools like GPT UiO do not provide a reliable source base and cannot search for sources. Evaluate the answers you receive critically and verify the information and claims by checking credible sources. Additionally, you should always look up your own sources and list/reference the sources you've used.
The answers you get from using artificial intelligence-based tools might be influenced by systematic biases in the data they are trained on. To avoid such bias, you should scrutinise the answers from the tool so that perspectives are not based on biases arising from, for instance, incomplete data on gender, minorities, or demographic factors.
Text-generating AI tools, for example, might produce immoral or objectionable content.
Legal Guidelines
Please make yourself familiar with Legal guidelines for the use of Artificial Intelligence (AI) at the University of Oslo before incorporating AI into your work or studying.