Risky-Bench: Evaluating Agentic Safety Risks in Real-World Life-Assist Deployments

arXiv:2602.03100 · cs.AI · Submitted 2026-02-03 · Read on arXiv

cs.AI

Submitted: 2026-02-03

Updated: 2026-10-11

Code: https://github.com/SophieZheng998/Risky-Bench

Terminology

• Adaptability — 1× in this paper · explained in 1 episode
• Large Language Models (LLMs) — 1× in this paper · explained in 6 episodes

Sources

• SafeSearch: Automated Red-Teaming of LLM-Based Search Agents
• Oyster-I: Beyond Refusal -- Constructive Safety Alignment for Responsible Language Models
• VitaBench: Benchmarking LLM Agents with Versatile Interactive Tasks in Real-world Applications
• Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety
• MobileSafetyBench: Evaluating Safety of Autonomous Agents in Mobile Device Control
• AgentSafe: Safeguarding Large Language Model-based Multi-agent Systems via Hierarchical Data Management
• UserRL: Training Interactive User-Centric Agent via Reinforcement Learning
• WebResearcher: Unleashing unbounded reasoning capability in Long-Horizon Agents
• PropensityBench: Evaluating Latent Safety Risks in Large Language Models via an Agentic Approach
• Your Agent May Misevolve: Emergent Risks in Self-evolving LLM Agents
• \texttt{R$^\textbf{2}$AI}: Towards Resistant and Resilient AI in an Evolving World
• A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deployment
• Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memory
• SafeSearch: Do Not Trade Safety for Utility in LLM Search Agents
• Agent-SafetyBench: Evaluating the Safety of LLM Agents
• DeepResearcher: Scaling Deep Research via Reinforcement Learning in Real-world Environments

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems