Swarming systems, such as for example multi-drone networks, excel at cooperative tasks like monitoring, surveillance, or disaster assistance in critical environments, where autonomous agents make decentralized decisions in order to fulfill team-level objectives in a robust and efficient manner. Unfortunately, team-level coordinated strategies in the wild are vulnerable to data poisoning attacks, resulting in either inaccurate coordination or adversarial behavior among the agents. To address this challenge, we contribute a framework that investigates the effects of such data poisoning attacks, using explainable AI methods. We model the interaction among agents using evolutionary intelligence, where an optimal coalition strategically emerges to perform coordinated tasks. Then, through a rigorous evaluation, the swarm model is systematically poisoned using data manipulation attacks. We showcase the applicability of explainable AI methods to quantify the effects of poisoning on the team strategy and extract footprint characterizations that enablediagnosing. Our findings indicate that when the model is poisoned above 10%, non-optimal strategies resulting in inefficient cooperation can be identified
Asadi, M, Radulescu, R & Nowe, A 2025, Explainable AI Based Diagnosis of Poisoning Attacks in Evolutionary Swarms. in G Ochoa (ed.), GECCO 2025 Companion - Proceedings of the 2025 Genetic and Evolutionary Computation Conference Companion. pp. 251-254, The Genetic and Evolutionary Computation Conference (GECCO 2025), Málaga, Spain, 14/07/25. https://doi.org/10.1145/3712255.3726576
Asadi, M., Radulescu, R., & Nowe, A. (2025). Explainable AI Based Diagnosis of Poisoning Attacks in Evolutionary Swarms. In G. Ochoa (Ed.), GECCO 2025 Companion - Proceedings of the 2025 Genetic and Evolutionary Computation Conference Companion (pp. 251-254) https://doi.org/10.1145/3712255.3726576
@inproceedings{5ce384325fe6403bae1d7d6532bce8a3,
title = "Explainable AI Based Diagnosis of Poisoning Attacks in Evolutionary Swarms",
abstract = "Swarming systems, such as for example multi-drone networks, excel at cooperative tasks like monitoring, surveillance, or disaster assistance in critical environments, where autonomous agents make decentralized decisions in order to fulfill team-level objectives in a robust and efficient manner. Unfortunately, team-level coordinated strategies in the wild are vulnerable to data poisoning attacks, resulting in either inaccurate coordination or adversarial behavior among the agents. To address this challenge, we contribute a framework that investigates the effects of such data poisoning attacks, using explainable AI methods. We model the interaction among agents using evolutionary intelligence, where an optimal coalition strategically emerges to perform coordinated tasks. Then, through a rigorous evaluation, the swarm model is systematically poisoned using data manipulation attacks. We showcase the applicability of explainable AI methods to quantify the effects of poisoning on the team strategy and extract footprint characterizations that enablediagnosing. Our findings indicate that when the model is poisoned above 10%, non-optimal strategies resulting in inefficient cooperation can be identified",
author = "Mehrdad Asadi and Roxana Radulescu and Ann Nowe",
note = "Publisher Copyright: {\textcopyright} 2025 Copyright held by the owner/author(s).; The Genetic and Evolutionary Computation Conference (GECCO 2025) ; Conference date: 14-07-2025 Through 18-07-2025",
year = "2025",
month = aug,
day = "11",
doi = "10.1145/3712255.3726576",
language = "English",
pages = "251--254",
editor = "Gabriela Ochoa",
booktitle = "GECCO 2025 Companion - Proceedings of the 2025 Genetic and Evolutionary Computation Conference Companion",
}