Newest papers

Adaptive Cyber-Defense for Evolving Attacker Capabilities using Reinforcement Learning and Behavior Trees

Nicholas Potteiger, Ankita Samaddar, Xenofon Koutsoukos · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.25420 · cited by 0 works (OpenAlex) · open-access PDF

Japanese

エンタープライズのコンピュータネットワークを守るために、強化学習(RL)を使用して訓練された自律型サイバー防衛エージェントはますます採用されています。 しかし、既存のアプローチは、攻撃者の能力が進化し、新たな攻撃戦略を可能にするにつれて脆弱になる可能性があります。この研究では、CAGE Challenge Scenario 2(CC2)で学んだ防衛方針を適応する問題を研究しています。この問題に対処するために、攻撃者の行動に基づく効果的な攻撃戦略は、当初は未知であり、したがって防衛適応が起こる前に攻撃者の方針を学ばなければなりません。これに対処するために、攻撃者の行動に対応するために、攻撃方針(ARLBT)を用いて、報酬行動

Machine translation (Meta M2M100); the English original follows.

Abstract

Autonomous cyber-defense agents trained using reinforcement learning (RL) are increasingly adopted for defending enterprise computer networks. However, existing approaches may become brittle as attacker capabilities evolve and enable new attack strategies. In this work, we study the problem of adapting defender policies learned in CAGE Challenge Scenario 2 (CC2), a benchmark enterprise network defense scenario, to an extended network scenario with new attacker capabilities represented by new attacker actions. Effective attack strategies that leverage new actions are initially unknown, and therefore attacker policies must be learned before defender adaptation can occur. To address this challenge, we develop adversarial RL (ARL) with behavior trees (ARLBT), which uses reward behavior trees (RBTs) to define progression-based reward functions and condition attacker policies on RBT execution. The proposed RBT incorporates the new attacker actions into structured attack strategies while leveraging parallel execution to generate rewards across multiple network nodes simultaneously. We evaluate ARLBT for learning attacker policies against two baseline defender policies across network scenarios with different sets of new attacker actions and compare against multiple attacker baselines. The results demonstrate that ARLBT improves attacker success compared to ARL using the original CC2 reward function and that conditioning policies on RBT execution can improve sample efficiency. We further show that defender policies can be adapted against the learned attackers, improving robustness to attacks that leverage new attacker actions.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Reinforcement learning, Computer science, Adversarial system, Leverage (statistics), Computer security, Robustness (evolution)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange