MetaPoison: Practical General-purpose Clean-label Data Poisoning

TL;DR

MetaPoison introduces a practical, general-purpose method for clean-label data poisoning that effectively fools neural networks across various models and training scenarios, including real-world API systems.

Contribution

It presents MetaPoison, a novel first-order meta-learning approach that approximates bilevel optimization for effective, transferable, and versatile data poisoning attacks on deep neural networks.

Findings

MetaPoison outperforms previous clean-label poisoning methods.

Poisoned data transfer across different models and training settings.

Successful real-world poisoning of Google Cloud AutoML models.

Abstract

Data poisoning -- the process by which an attacker takes control of a model by making imperceptible changes to a subset of the training data -- is an emerging threat in the context of neural networks. Existing attacks for data poisoning neural networks have relied on hand-crafted heuristics, because solving the poisoning problem directly via bilevel optimization is generally thought of as intractable for deep models. We propose MetaPoison, a first-order method that approximates the bilevel problem via meta-learning and crafts poisons that fool neural networks. MetaPoison is effective: it outperforms previous clean-label poisoning methods by a large margin. MetaPoison is robust: poisoned data made for one model transfer to a variety of victim models with unknown training settings and architectures. MetaPoison is general-purpose, it works not only in fine-tuning scenarios, but also for end-to-end training from scratch, which till now hasn't been feasible for clean-label attacks with deep nets. MetaPoison can achieve arbitrary adversary goals -- like using poisons of one class to make a target image don the label of another arbitrarily chosen class. Finally, MetaPoison works in the real-world. We demonstrate for the first time successful data poisoning of models trained on the black-box Google Cloud AutoML API. Code and premade poisons are provided at https://github.com/wronnyhuang/metapoison