{
  "id": 7124,
  "url": "https://arxiv.org/abs/2603.16969v2",
  "title": "DeepStage: Learning Autonomous Defense Policies Against Multi-Stage APT Campaigns",
  "summary": "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 attac",
  "authors": "Trung V. Phan, Tri Gia Nguyen, Thomas Bauschert",
  "category": "research",
  "topics": "military-security,environment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-03-17T09:46:11.000Z",
  "fetched_at": "2026-07-14T16:32:59.166Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/7124",
  "original_url": "https://arxiv.org/abs/2603.16969v2",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}