Experimental Attacks on Quantum Computing and Quantum Machine Learning

Abstract

2025 marks the UN International Year of Quantum Science and Technology. We expect this emerging technology to enter application in real world industry use cases in the next few years. The current focus in research and industry lies on the nominal function of quantum computing to show its “usefulness” of the technology. Hence, one should also start considering aspects of cybersecurity to prepare for currently known attacks when quantum computing is ready to be used in industry.This work provides an overview on quantum computing (QC) with the underlying used quantum mechanical phenomena. We set special highlight on quantum machine learning (QML) as it is expected to have higher expressiveness and shorter training times. The quantum machine learning lifecycle will be presented in detail with the current state of the art of theory and research of cybersecurity attacks.We aim to use this knowledge and start to construct simple experiments as proof-of-concepts for attacking assets in the CIA triad. These proof-of-concepts will set the base for attacks on more relevant use cases of QML models, leading to thoughts and experiments for countermeasures to be set up and put in place. The conclusion will consist of future work and the currently not in detail investigated sections of the lifecycle.