aiqbench analysis
This research introduces FedMAST, a defense mechanism for federated learning that detects stealthy backdoor attacks by analyzing residual structural traces in client updates rather than relying on isolated anomaly signals. By combining squeeze-pair coherence scoring with signed spectral-drift tracking, the method effectively suppresses adversarial influence while maintaining high accuracy against diverse poisoning strategies. The findings are critical for securing distributed AI training pipelines where data privacy constraints prevent centralized oversight.
Backdoors Leave Structural Traces: FedMAST for Backdoor Detection and Containment in Federated Learning
arXiv:2609.27760v1 Announce Type: cross Abstract: Federated learning enables distributed training of a shared model without requiring clients to share their raw data. However, its reliance on the integrity of the client-submitted updates exposes the global model to stealthy backdoor poisoning. Although existing defenses often inspect isolated evidence sources, stealth-constrained attacks can adapt to these signals. In this paper, we show that such attacks can suppress isolated anomaly signals, but their poisoned updates still leave residual structural traces. We propose FedMAST, a Federated Multi-Axis Structural Tracing defense for backdoor detection in federated learning. FedMAST scores client updates using complementary structural, spectral, and historical evidence and then applies tiered filtering and round-level containment to limit adversarial influence. To capture traces that isolated signals may miss, FedMAST uses squeeze-pair coherence scoring to expose coupled feature distortions and signed spectral-drift tracking to reveal persistent directional changes over time. Across six federated backdoor attacks, namely Constrain-and-Scale, Neurotoxin, BC-Layers, LGA, DBA, and 3DFed, FedMAST achieves lower ASR than baseline defenses in all nine evaluated attack--defense comparisons. Across the complete 200-round runs, it attains an average ASR of 1.51% while maintaining 94.84% average main-task accuracy. Under the method-aware CovertLayers attack, FedAvg, MultiKrum, AlignIns, and FLAME yield full-run ASRs of 100.00%, 99.67%, 99.53%, and 32.84%, respectively. FedMAST achieves the lowest ASR among all evaluated methods, reducing it to 1.53% while maintaining 92.26% main-task accuracy.
Abstract:Federated learning enables distributed training of a shared model without requiring clients to share their raw data. However, its reliance on the integrity of the client-submitted updates exposes the global model to stealthy backdoor poisoning. Although existing defenses often inspect isolated evidence sources, stealth-constrained attacks can adapt to these signals. In this paper, we show that such attacks can suppress isolated anomaly signals, but their poisoned updates still leave residual structural traces. We propose FedMAST, a Federated Multi-Axis Structural Tracing defense for backdoor detection in federated learning. FedMAST scores client updates using complementary structural, spectral, and historical evidence and then applies tiered filtering and round-level containment to limit adversarial influence. To capture traces that isolated signals may miss, FedMAST uses squeeze-pair coherence scoring to expose coupled feature distortions and signed spectral-drift tracking to reveal persistent directional changes over time. Across six federated backdoor attacks, namely Constrain-and-Scale, Neurotoxin, BC-Layers, LGA, DBA, and 3DFed, FedMAST achieves lower ASR than baseline defenses in all nine evaluated attack--defense comparisons. Across the complete 200-round runs, it attains an average ASR of 1.51% while maintaining 94.84% average main-task accuracy. Under the method-aware CovertLayers attack, FedAvg, MultiKrum, AlignIns, and FLAME yield full-run ASRs of 100.00%, 99.67%, 99.53%, and 32.84%, respectively. FedMAST achieves the lowest ASR among all evaluated methods, reducing it to 1.53% while maintaining 92.26% main-task accuracy.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2609.27760 [cs.LG]
(or arXiv:2609.27760v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.27760 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 318 words
Related coverage
An unexamined cause of the OpenAI Hugging Face hacking incident: its binary performance metric
LessWrong · Sep 23, 2026
Informe de riesgos de la IA de frontera (febrero–marzo de 2026)
METR · May 19, 2026
Frontier AI Risk Report (February–March 2026)
METR · May 19, 2026
Frontier Risk Report (February to March 2026)
METR · May 19, 2026