Assessing the threat levels of misinformation campaigns
Led by Christina Pöpper
This research aims to build a Machine Learning based framework that assesses the threat level of misinformation campaigns for societies. Misinformation campaigns are a modern form of information warfare. The research focuses on assessing the threat levels to prioritize the attention of mitigating agencies on campaigns that may pose more significant threats to civil society. This is a timely goal in the current political atmosphere and the age of information flooding, where misinformation is targeting key global vaccination campaigns such as that against COVID-19.
Stealthy attacks on autonomous vehicle-based control systems and their defenses
Led by Muhammad Shafique
This research aims to investigate attacks and defenses for the machine learning modules in connected and automated vehicles (CAV). With increased automation comes increased vulnerability to cyber-attacks that can hack a vehicle’s electronic systems. Researchers have demonstrated an ability to take over a vehicle’s electronic systems and cause crashes. The research focuses on a new type of attacks on the deep neural networks of CAV, the so-called Backdoored Neural Networks, that only behave maliciously when triggered by specific inputs and on the relevant mitigation strategies.