{
  "id": 5580,
  "url": "https://arxiv.org/abs/2604.18847v2",
  "title": "Human-Guided Harm Recovery for Computer Use Agents",
  "summary": "As LM agents gain the ability to execute actions on real computer systems, we need ways to not only prevent harmful actions at scale but also effectively remediate harm when prevention fails. We formalize a solution to this neglected challenge in post-execution safeguards as harm recovery: the problem of optimally steering an agent from a harmful state back to a safe one in alignment with human preferences. We ground preference-aligned recovery through a formative user study that identifies valu",
  "authors": "Christy Li, Sky CH-Wang, Andi Peng, Andreea Bobu",
  "category": "research",
  "topics": "safety-alignment,agents-autonomy",
  "orgs": null,
  "regions": null,
  "published_at": "2026-04-20T21:12:40.000Z",
  "fetched_at": "2026-07-14T16:31:53.163Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/5580",
  "original_url": "https://arxiv.org/abs/2604.18847v2",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}