Modeling and Security Analysis of Attacks on Machine Learning Systems

Published

March 9, 2026

Author(s)

Anoop Singhal, Qingtian Zou

Abstract

Machine learning (ML) systems are more and more widely deployed. At the same time, ML systems also introduce new security threats that are not seen in traditional software and network systems. For example, data poisoning and adversarial examples attacks generate incorrect output; membership and denial of service attacks can also exploit the new attack surface which waste computation resources. However, existing security causality analysis frameworks fail to sufficiently handle the new causality relationships, notions and elements introduced by ML systems. In this paper, we present a new way to model causality relationships in ML systems for security analysis. We present our model with a running example and discuss how our model can be used for security analysis and risk mitigation.

Proceedings Title

2025 IEEE 7th International Conference on Cognitive Machine Intelligence (CogMI)

Citation

<i>2025 IEEE 7th International Conference on Cognitive Machine Intelligence (CogMI)</i> (2025 IEEE 7th International Conference on Cognitive Machine Intelligence (CogMI), Pittsburgh, PA, US, 12-NOV-25 to 14-NOV-25)

Conference Dates

2025-11-12T00:00:00

Conference Location

Pittsburgh, PA, US

Conference Title

2025 IEEE 7th International Conference on Cognitive Machine Intelligence (CogMI)

Pub Type

Conferences

Download Paper

https://doi.org/10.1109/CogMI67134.2025.00061

Local Download

Keywords

machine learning, system security, causality modeling, attack graphs

Citation

Singhal, A. and Zou, Q. (2026), Modeling and Security Analysis of Attacks on Machine Learning Systems, 2025 IEEE 7th International Conference on Cognitive Machine Intelligence (CogMI), Pittsburgh, PA, US, [online], https://doi.org/10.1109/CogMI67134.2025.00061, https://nps-ext.nist.gov/nps/nps_public_api/api/Publication/file/958427 (Accessed November 13, 2026)

Additional citation formats