Guy Amit (Ben-Gurion University), Moshe Levy (Ben-Gurion University), Yisroel Mirsky (Ben-Gurion University)

Deep neural networks are normally executed in the forward direction. However, in this work, we identify a vulnerability that enables models to be trained in both directions and on different tasks. Adversaries can exploit this capability to hide rogue models within seemingly legitimate models. In addition, in this work we show that neural networks can be taught to systematically memorize and retrieve specific samples from datasets. Together, these findings expose a novel method in which adversaries can exfiltrate datasets from protected learning environments under the guise of legitimate models.

We focus on the data exfiltration attack and show that modern architectures can be used to secretly exfiltrate tens of thousands of samples with high fidelity, high enough to compromise data privacy and even train new models. Moreover, to mitigate this threat we propose a novel approach for detecting infected models.

View More Papers

Eavesdropping on Black-box Mobile Devices via Audio Amplifier's EMR

Huiling Chen (College of Computer Science and Electronic Engineering, Hunan University, Changsha, China), Wenqiang Jin (College of Computer Science and Electronic Engineering, Hunan University, Changsha, China), Yupeng Hu (College of Computer Science and Electronic Engineering, Hunan University, Changsha, China), Zhenyu Ning (College of Computer Science and Electronic Engineering, Hunan University, Changsha, China), Kenli Li (College…

Read More

On the Feasibility of CubeSats Application Sandboxing for Space...

Gabriele Marra (CISPA Helmholtz Center for Information Security), Ulysse Planta (CISPA Helmholtz Center for Information Security and Saarbrücken Graduate School of Computer Science), Philipp Wüstenberg (Chair of Space Technology, Technische Universität Berlin), Ali Abbasi (CISPA Helmholtz Center for Information Security)

Read More

Facilitating Threat Modeling by Leveraging Large Language Models

Isra Elsharef, Zhen Zeng (University of Wisconsin-Milwaukee), Zhongshu Gu (IBM Research)

Read More