Skip to Main Content
Turku University Library

Library Guide to Artificial Intelligence

AI Guide for Information Seekers

This guide is designed to support information seekers in the responsible use of artificial intelligence. It provides information about AI tools offered by the library, the potential of AI in information searching, and tips for ethical and safe use. The aim is to help users make considered and responsible use of AI when searching for information.

An increasing number of databases provided by the library now include AI-assisted features, such as natural language search, AI assistants, and content structuring tools (e.g. summaries and keyword lists). These features support reading, understanding information, and grasping broader contexts, offering a new approach to information searching. AI-based tools do not replace traditional search methods, but rather complement them by providing alternative ways to approach research questions and organise large bodies of information.

AI-assisted information searching should always begin with an assessment of your own information needs and the content available in the tools you plan to use. It is important to understand what material each tool draws upon, as the quality of search results depends on how well the underlying content matches your research topic or information need.

This AI guide evolves alongside the technology: its content is updated regularly to reflect developments in AI and changing information-seeking practices.

For more information about other AI tools provided by the university, please refer to the university's intranet page on AI.

AI in Information Seeking

Artificial intelligence – meaning the ability of computer systems to perform tasks that normally require human intelligence – is widely used in everyday life, for example in recommendation systems and language translation. From the perspective of information seeking, AI enables new ways of searching, organizing, and producing information.

In traditional information seeking, the user controls the process: they select an appropriate database and search terms, combine the terms into a search query using Boolean operators, apply filters, and choose the most relevant results themselves. The results are publications or references whose origin the user can verify. You can find more information about traditional information searching in the library’s Basics of Information Seeking guide.

Generative AI does not search information directly from external sources, but instead produces answers based on its training data and language model structures. In practice, the language model predicts probable words and builds answers from them. Sometimes generative AI may hallucinate – that is, generate responses that sound reliable but are not based on actual knowledge. Using generative AI resembles a conversation more than traditional information seeking: the user poses a question or a prompt, and the AI generates a response based on its content and context. Examples of such tools include Copilot and ChatGPT when used without web search.

RAG search (Retrieval-Augmented Generation) combines traditional information retrieval with generative AI. In this model, the AI first searches information from external sources – such as a specific database – and then provides the relevant results to the language model, which uses them to generate a response. This way, answers can be more up to date and the sources visible. However, it is important to note that RAG models can also produce inaccurate answers: even if the information is searched from a reliable source, the language model may phrase the response in a way that no longer reflects the original content. Examples of RAG-based AI tools integrated into databases include Volter’s AI assistant or Scopus AI.

Artificial Intelligence at the University

University of Turku has established guidelines for the use of artificial intelligence in teaching, studying, and research. These guidelines provide direction for the responsible use of AI. In addition to following the university-wide policies, students should check at the course level whether the use of AI is permitted, how it may be used, and how its use should be reported.

The university has compiled support materials for AI use on the intranet site AI at the University. The site includes information on responsible use of AI and lists all AI solutions provided by the university.