Artificial Intelligence (AI) has been around as a concept and technology for many years now and you will have used tools which utilise AI, in the form of machine learning, in your studies without realising. However, the more recent development of Generative AI (GenAI) has had a far greater impact upon education and how you find and engage with information.
From a library perspective we are keen to ensure that your use of GenAI tools is as effective and appropriate as it can be. But there are so many tools (with new ones being launched or developed on a daily basis) that it is very difficult for us to recommend or demonstrate any specific tools that would support your research skills.
As such we believe that a better approach is to look at the commonalities between these tools and what skills you need to develop to use any tool effectively. The resulting skillset we have identified is not that different from what we already support you in developing for using more traditional search tools, information and referencing - it's just a shift in focus.
On this page you will find advice in relation to three key areas of research - Find, Check and Use - and how you can adapt your existing skillset to be able to approach the use of Generative AI tools for searching, evaluating and appropriately using information.
GenAI can certainly assist in finding resources. Previously students would conduct a search strategy, using keywords and refiners, now with GenAI, the search strategy is conducted in a slightly different way with prompt engineering. Prompts are used to ‘ask’ GenAI to deliver results for resources
To help students achieve this we have developed a prompt framework:
Help AI help you. Ask as clearly and directly as possible. By being specific and clearly defining exactly what it is you're looking for, the results you receive will be much more relevant and accurate.
Less is more. Try to keep your search simple by focusing on a few key words. Furthermore, try to avoid jargon, slang or filler words, which may create unnecessary confusion.
Context is everything. Consider the background and parameters of your search. This will provide a helpful foundation for your searching and ensure you don't drift from the main focus.
It is always important that the information you use in your assignments is factual, reliable, evidence-based and comes from a place of academic or professional expertise. To establish this in relation to any information source we recommend the TRYangle Approach which encourages you to approach information from 3 interlinked areas of understanding. For each of these 3 areas, here are some particular considerations you may wish to make in relation to AI generated content:
Often with AI generated content it is not possible to establish who the author is, how current the information is and where it has come from, as it is actually the merging together of a range of different sources. Some tools will cite the sources used and so you can evaluate whether these are reliable using the principles of lateral searching. In cases where the source isn’t given, you would not want to be using the generated information in your assignments. Instead you should only be using it as a starting point to identify key concepts and then searching for more reliable sources relating to these.
It is important to understand how the information has arrived at you. GenAI tools produce information by studying pools of data that have been selected to train them upon. From this data the tool identifies patterns and relationships. When you enter a prompt it then forms content using those patterns to predict a relevant sequence of words (or pixels for images). If there are gaps in the output the tool will fill these with information that seems to fit the pattern, regardless of whether it is factual – this is often referred to as “hallucinations”.
Because the output of GenAI tools is pooled from across a range of sources it will only ever represent the dominant thoughts or ideas, there is much less space for diverse voices. This also means that it will promote biases inherent within the training data. With the output also being specific to your prompt there is also a higher danger of confirmation bias, with the potential for the information to echo your own views and opinions.
Check your assignment brief to determine if the use of AI tools is permitted, and to understand the guidelines on what you are allowed and not allowed to do.
Before incorporating AI outputs in your work, think if they are really supporting your learning, and help you develop your academic skills. AI tools can help you with generating ideas, suggesting structures or summarising texts, but you need to think if it is appropriate for your assignment to copy-paste text generated from a prompt directly into it. It might be better to use your research skills to follow through to the original sources and not use what AI produced.
Always cross-check AI generated outputs against established sources to verify accuracy and identify inaccurate information.
Avoid becoming overly reliant on AI tools, as this can hinder your own learning and development of critical thinking skills.
AI tools may not be suitable for all types of academic tasks, particularly those that require critical thinking, analysis, or creative writing.
As with all information sources, you should always acknowledge any content that you have used to inform your assignments. In the same way that you would offer an in-text citation and entry in your reference list for a book, journal article or web page, you should acknowledge the use of information generated by an AI tool too.
Please see guidance on our Harvard Solent Referencing Guide on how to correctly cite AI generated content:
Inappropriate use without acknowledging the use of AI tools might lead to academic misconduct. For further information, please see Student Academic Misconduct.
We have also put all of the information on this page into a Research Skills & AI video, which is also embedded below, for you to explore these concepts and guidance in a more visual manner.