There are many AI tools specifically designed for academic literature research. Here you find an overview of these tools as well as information on the subjects and use cases for which they are (not yet) suited.
Further down below, you also find an overview of many other AI tools that can be useful for academic research.
The University Library offers workshops, consultations, and self-study courses on "Literature Research with AI".
Please note when using the tools listed here:
Practically all of them are problematic with respect to data privacy. Please do not enter any personal data of third parties and carefully consider to what extent you want to disclose your own data.
Please do not upload any copyrighted documents to these tools.
Please make sure that your use of these tools is in line with the examination regulations of your department.
Literature Research with AI - Tools and Introduction
AI tools for academic literature research can be divided into "Finders" and "Connectors".
Here you will find a detailed list of the most important Finders and Connectors.
"Finders" work in a similar way to library catalogues: you enter a keyword, a phrase or - even better with many tools - a complete question. The tool then shows you a list of results.
"Connectors" are based on a publication that you have already found. You enter part of this publication’s metadata (preferably the DOI) into the tool in order to find related literature.
"Finders" that, in our experience, already deliver quite good results are
You can find more information on the tools mentioned here, in our training courses and consultations as well as in our self-study course. You can find many more AI tools for academic work here.
At the moment, the "Finders" are mainly suitable for research in the natural sciences and medicine, less so for the humanities and social sciences. These tools mostly find journal articles (i.e. not monographs) in English - especially if they have a DOI and have been published Open Access. If these criteria usually apply to the literature in your subject area, the "Finders" should be able to provide you with rather good results.
For the humanities and social sciences, we recommend trying out the "Connectors".
Please also note: Chatbots such as OpenAI's ChatGPT, Google's Gemini, Microsoft's Copilot, Anthropic's ClaudeAI, Mistral's Le Chat etc. are currently not yet suitable for literature research, as they mainly provide hallucinated (i.e. invented) literature references in response to queries of this kind.
Furthermore, in our experience, Perplexity AI is currently only suitable for searching for 'everyday information', but not for scientific research.
On request, we are also happy to offer additional workshops (in German or English) for groups of five or more participants. If you are interested in this offer, just contact us at informationspam prevention@ub.uni-tuebingen.de.
