Cover Image

Medical large language models are susceptible to targeted misinformation attacks

Large language models (LLMs) have broad medical knowledge and can reason about medical information across many domains, holding promising potential for diverse medical applications in the near future. In this study, we demonstrate a concerning vulnerability of LLMs in medicine. Through targeted mani...

Full description

Saved in:
Bibliographic Details
Main Authors: Han, Tianyu (Author) , Nebelung, Sven (Author) , Khader, Firas (Author) , Wang, Tianci (Author) , Müller-Franzes, Gustav (Author) , Kuhl, Christiane (Author) , Försch, Sebastian (Author) , Kleesiek, Jens Philipp (Author) , Haarburger, Christoph (Author) , Bressem, Keno K. (Author) , Kather, Jakob Nikolas (Author) , Truhn, Daniel (Author)
Format: Article (Journal)
Language:English
Published: 23 October 2024
In: npj digital medicine
Year: 2024, Volume: 7, Pages: 1-9
ISSN:2398-6352
DOI:10.1038/s41746-024-01282-7
Online Access:Verlag, kostenfrei, Volltext: https://doi.org/10.1038/s41746-024-01282-7
Verlag, kostenfrei, Volltext: https://www.nature.com/articles/s41746-024-01282-7
Get full text
Author Notes:Tianyu Han, Sven Nebelung, Firas Khader, Tianci Wang, Gustav Müller-Franzes, Christiane Kuhl, Sebastian Försch, Jens Kleesiek, Christoph Haarburger, Keno K. Bressem, Jakob Nikolas Kather & Daniel Truhn
Description
Summary:Large language models (LLMs) have broad medical knowledge and can reason about medical information across many domains, holding promising potential for diverse medical applications in the near future. In this study, we demonstrate a concerning vulnerability of LLMs in medicine. Through targeted manipulation of just 1.1% of the weights of the LLM, we can deliberately inject incorrect biomedical facts. The erroneous information is then propagated in the model’s output while maintaining performance on other biomedical tasks. We validate our findings in a set of 1025 incorrect biomedical facts. This peculiar susceptibility raises serious security and trustworthiness concerns for the application of LLMs in healthcare settings. It accentuates the need for robust protective measures, thorough verification mechanisms, and stringent management of access to these models, ensuring their reliable and safe use in medical practice.
Item Description:Gesehen am 23.04.2025
Physical Description:Online Resource
ISSN:2398-6352
DOI:10.1038/s41746-024-01282-7

Similar Items