RL-based safety alignment with an execute-refuse-verify policy improves reported threat resistance for tool-using agents, but utility preservation is not consistently demonstrated.
AI agents under threat: A survey of key security challenges and future pathways
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.AI 1years
2025 1verdicts
REJECT 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Agent Safety Alignment via Reinforcement Learning
RL-based safety alignment with an execute-refuse-verify policy improves reported threat resistance for tool-using agents, but utility preservation is not consistently demonstrated.