Widen The Backdoor To Let More Attackers In
Siddhartha Datta, Giulio Lovisotto, Ivan Martinovic, Nigel Shadbolt

TL;DR
This paper explores multi-agent backdoor attacks in collaborative learning, revealing that more attackers reduce individual success rates and proposing defenses by increasing attackers artificially and removing attacker data during inference.
Contribution
It uncovers the backfiring effect in multi-agent backdoor attacks and proposes two defenses leveraging this phenomenon to improve model robustness.
Findings
Increasing attackers decreases attack success rate
Artificially augmenting attackers reduces attack effectiveness
Removing attacker data during inference enhances robustness
Abstract
As collaborative learning and the outsourcing of data collection become more common, malicious actors (or agents) which attempt to manipulate the learning process face an additional obstacle as they compete with each other. In backdoor attacks, where an adversary attempts to poison a model by introducing malicious samples into the training data, adversaries have to consider that the presence of additional backdoor attackers may hamper the success of their own backdoor. In this paper, we investigate the scenario of a multi-agent backdoor attack, where multiple non-colluding attackers craft and insert triggered samples in a shared dataset which is used by a model (a defender) to learn a task. We discover a clear backfiring phenomenon: increasing the number of attackers shrinks each attacker's attack success rate (ASR). We then exploit this phenomenon to minimize the collective ASR of…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdversarial Robustness in Machine Learning · Anomaly Detection Techniques and Applications · Network Security and Intrusion Detection
