Human-Guided Harm Recovery for Computer Use Agents
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
Record details
Published: 20 April 2026
Source: arXiv
Category: Research
Topics: Safety & alignment · Agents & autonomy
Retrieved: 14 July 2026
Related evidence
These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.
SDOF: Taming the Alignment Tax in Multi-Agent Orchestration with State-Constrained Dispatch
arXiv · 20 April 2026
Four-Axis Decision Alignment for Long-Horizon Enterprise AI Agents
arXiv · 21 April 2026
Taming Actor-Observer Asymmetry in Agents via Dialectical Alignment
arXiv · 21 April 2026
Reinforcing privacy reasoning in LLMs via normative simulacra from fiction
arXiv · 21 April 2026
Semantic Prompting: Agentic Incremental Narrative Refinement through Spatial Semantic Interaction
arXiv · 21 April 2026
Waking Up Blind: Cold-Start Optimization of Supervision-Free Agentic Trajectories for Grounded Visual Perception
arXiv · 19 April 2026
How to cite this record
ethics.ai (20 April 2026), “Human-Guided Harm Recovery for Computer Use Agents,” evidence record 5580, https://ethics.ai/record/5580 (originally published by arXiv).
Use and limitations
This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.