{
  "id": 12987,
  "url": "https://arxiv.org/abs/2607.19829v1",
  "title": "DARWIN: Evolving Jailbreak Adversary and Guardrail for LLM Safety Evaluation and Protection",
  "summary": "Most existing LLM safety evaluation and defense methods follow a static formulation: jailbreak vulnerabilities are evaluated with fixed attack methods, and guardrails are trained on fixed malicious prompt datasets. However, real-world adversaries continuously evolve their capabilities and expand the attack space. To address this challenge, we propose DARWIN, an evolutionary attack-defense framework that formulates jailbreaking as an open-ended evolution process and continuously updates guardrail",
  "authors": "Weiwei Qi, Zefeng Wu, Zhilin Guo, Tianhang Zheng, Chaochao Lu, Liang He, Zhan Qin, Kui Ren",
  "category": "research",
  "topics": "safety-alignment,military-security",
  "orgs": null,
  "regions": null,
  "published_at": "2026-07-22T07:08:59.000Z",
  "fetched_at": "2026-07-23T05:10:49.458Z",
  "source_slug": "x-arxiv-red-teaming-query",
  "source_name": "arXiv red teaming query",
  "source_homepage": "https://arxiv.org/a/redteam",
  "ethics_ai_record_url": "https://ethics.ai/record/12987",
  "original_url": "https://arxiv.org/abs/2607.19829v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}