Accumulative Poisoning Attacks on Real-time Data

Tianyu Pang, Xiao Yang, Yinpeng Dong, Hang Su, Jun Zhu

Keywords: Poisoning attacks, Real-time data

TL;DR: We propose accumulative poisoning attacks, which can significantly break down the model by a single update step after an accumulative phase.

Abstract: Collecting training data from untrusted sources exposes machine learning services to poisoning adversaries, who maliciously manipulate training data to degrade the model accuracy. When trained on offline datasets, poisoning adversaries have to inject the poisoned data in advance before training, and the order of feeding these poisoned batches into the model is stochastic. In contrast, practical systems are more usually trained/fine-tuned on sequentially captured real-time data, in which case poisoning adversaries could dynamically poison each data batch according to the current model state. In this paper, we focus on the real-time settings and propose a new attacking strategy, which affiliates an accumulative phase with poisoning attacks to secretly (i.e., without affecting accuracy) magnify the destructive effect of a (poisoned) trigger batch. By mimicking online learning and federated learning on MNIST and CIFAR-10, we show that model accuracy significantly drops by a single update step on the trigger batch after the accumulative phase. Our work validates that a well-designed but straightforward attacking strategy can dramatically amplify the poisoning effects, with no need to explore complex techniques.

Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.

Code: https://github.com/ShawnXYang/AccumulativeAttack

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2106.09993/code)