Cost-Aware Hierarchical Multi-Agent Ransomware Detection and Family Attribution

arXiv 2609.04820 ↗First posted 4 Sep 2026cs.CR cs.AIcited by 0

On the map: LLM agents and tool use → Agent threats, guardrails and governance

Abstract (arXiv)

Sandbox execution and memory forensics are among the most constrained resources in malware triage. Static analysis can scale to millions of files, whereas dynamic and memory analysis require minutes of analyst controlled infrastructure for each sample. Despite this difference, multimodal ransomware detectors often apply every modality to every sample, causing analysis cost and time to verdict to increase linearly with sample volume even when static evidence is already sufficient for a decision. We present a cost aware Hierarchical Multi-Agent System that formulates evidence acquisition as a budgeted sequential decision problem. Specialist agents generate schema validated risk signals for each modality, domain controllers aggregate these signals, and a Meta-Orchestrator begins with static evidence and escalates to dynamic and memory evidence only when confidence is insufficient or agents within a controller disagree. An optional, bounded, locally hosted large language model reviewer can adjust a verdict by at most one tier but cannot replace the deterministic pipeline. Each decision is recorded with a complete provenance trace. In multiple runs over 12439 samples from 16 ransomware families and benign samples, the deterministic HMAS achieves F1 0.93 and macro F1 0.97, resolving 57.95% of cases using static evidence alone, 35.83% after adding dynamic evidence, and only 6.21% through the full pipeline. The average internal analysis cost is 6.65 units, compared with 12 for exhaustive analysis, representing a 44.6% reduction. Standalone leave-one-family-out testing further shows that accuracy on families held out during tuning falls to 0.26 to 0.64 outside the Benign and high support classes. We report these results alongside a cost sensitivity analysis, a partial leave one component out ablation, and a full scale comparison with learned early and late fusion and cascade baselines.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Agent threats, guardrails and governance, one of 16 directions in LLM agents and tool use. Builds on 17 of its 37 references that are in the map. First posted on arXiv 4 Sep 2026.

Builds on

17 of 37 references.

Closest ancestors · 1

Same direction, newest first

2024Teams of LLM Agents can Exploit Zero-Day Vulnerabilities2 Jun

Other references · 16

2017On Calibration of Modern Neural Networkscited by 4,860
2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,906
2017Selective Classification for Deep Neural Networkscited by 799
2021A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantificationcited by 777
2018TESSERACT: Eliminating Experimental Bias in Malware Classification across Space and Timecited by 108
2020Dos and Don'ts of Machine Learning in Computer Securitycited by 98
2024Large Language Models for Cyber Security: A Systematic Literature Reviewcited by 65
2018Exploring Adversarial Examples in Malware Detectioncited by 41
2017Classification with Costly Features using Deep Reinforcement Learningcited by 32
2024Hierarchical Multi-agent Reinforcement Learning for Cyber Network Defensecited by 16
2017Adaptive Classification for Prediction Under a Budgetcited by 15
2021A Survey on Ransomware: Evolution, Taxonomy, and Defense Solutionscited by 9

4 more

2023Prompt Engineering-assisted Malware Dynamic Analysis Using GPT-4cited by 7
2024SLIFER: Investigating Performance and Robustness of Malware Detection Pipelinescited by 4
2024Zero Day Ransomware Detection with Pulse: Function Classification with Transformer Models and Assembly Languagecited by 2
2026Multimodal Multi-Agent Ransomware Analysis Using AutoGencited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way