Two Can Play That Game: An Adversarial Evaluation of a Cyber-alert Inspection System

TL;DR

This paper evaluates a reinforcement learning-based cyber-alert inspection system by developing adversarial attack strategies, revealing its robustness, and proposing game-theoretic enhancements to improve defense against sophisticated attackers.

Contribution

It introduces an adversarial evaluation framework for RL-based cyber-defense, extends the model to a game-theoretic setting, and demonstrates how to enhance defender robustness using a double oracle approach.

Findings

Defender policy is robust to the best response attacker.

Game-theoretic defender policies can be robust against any adversarial policy.

The double oracle approach improves defender robustness against discovered attacker policies.

Abstract

Cyber-security is an important societal concern. Cyber-attacks have increased in numbers as well as in the extent of damage caused in every attack. Large organizations operate a Cyber Security Operation Center (CSOC), which form the first line of cyber-defense. The inspection of cyber-alerts is a critical part of CSOC operations. A recent work, in collaboration with Army Research Lab, USA proposed a reinforcement learning (RL) based approach to prevent the cyber-alert queue length from growing large and overwhelming the defender. Given the potential deployment of this approach to CSOCs run by US defense agencies, we perform a red team (adversarial) evaluation of this approach. Further, with the recent attacks on learning systems, it is even more important to test the limits of this RL approach. Towards that end, we learn an adversarial alert generation policy that is a best response to the defender inspection policy. Surprisingly, we find the defender policy to be quite robust to the best response of the attacker. In order to explain this observation, we extend the earlier RL model to a game model and show that there exists defender policies that can be robust against any adversarial policy. We also derive a competitive baseline from the game theory model and compare it to the RL approach. However, we go further to exploit assumptions made in the MDP in the RL model and discover an attacker policy that overwhelms the defender. We use a double oracle approach to retrain the defender with episodes from this discovered attacker policy. This made the defender robust to the discovered attacker policy and no further harmful attacker policies were discovered. Overall, the adversarial RL and double oracle approach in RL are general techniques that are applicable to other RL usage in adversarial environments.