# DeepStage: Learning Autonomous Defense Policies Against Multi-Stage APT Campaigns

arXiv: https://arxiv.org/abs/2603.16969 · HTML: https://mlsanity.com/paper/2603.16969/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Adversarial robustness and RL in finance, one of 15 directions in Deep reinforcement learning methods.
- Builds on 5 of its 19 references that are in the map.
- First posted on arXiv 17 Mar 2026.

## Placed in

- Deep reinforcement learning methods / [Adversarial robustness and RL in finance](https://mlsanity.com/deeprl/adversarial-robustness-and-rl-in-finance/)

## Builds on (most cited first, 5 of 5)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Deep Reinforcement Learning for Cyber Security](https://mlsanity.com/paper/1906.05799/) (2019)
- [Practical Whole-System Provenance Capture](https://mlsanity.com/paper/1711.05296/) (2017)
- [Dynamic SDN-based Radio Access Network Slicing with Deep Reinforcement Learning for URLLC and eMBB Services](https://mlsanity.com/paper/2202.06435/) (2022)
- [Learning the APT Kill Chain: Temporal Reasoning over Provenance Data for Attack Stage Estimation](https://mlsanity.com/paper/2603.07560/) (2026)

## Built on by (most cited first, 3 of 3)

- [DeepXplain: XAI-Guided Autonomous Defense Against Multi-Stage APT Campaigns](https://mlsanity.com/paper/2603.21296/) (2026)
- [DeepFaith: Evidence-Grounded LLMs for Faithful Incident Reporting in Multi-Stage APT Defense](https://mlsanity.com/paper/2607.24348/) (2026)
- [A Cyber Range Evaluation of Autonomous Network Incident Response Agents](https://mlsanity.com/paper/2609.16541/) (2026)

## Abstract (arXiv)

This paper presents DeepStage, a deep reinforcement learning (DRL) framework for adaptive and stage-aware defense against Advanced Persistent Threats (APTs). The enterprise environment is formulated as a partially observable Markov decision process (POMDP), in which host provenance and network telemetry are fused into unified provenance graphs. Building on our prior work (StageFinder), DeepStage employs a graph neural network encoder and an LSTM-based stage estimator to infer probabilistic attacker stages aligned with the MITRE ATT&CK framework. The resulting stage beliefs, together with graph embeddings, are used to guide a hierarchical Proximal Policy Optimization (PPO) agent that selects defense actions across monitoring, access control, containment, and remediation. Experiments in a realistic enterprise testbed with CALDERA-driven APT playbooks show that DeepStage achieves an average F1-score of 0.887 and a mitigation success rate of 84.7%, outperforming a risk-aware DRL baseline by 21.8% in F1-score and 16.2% in mitigation success. The results demonstrate effective stage-aware and cost-efficient autonomous cyber defense.