agent-risk: Risk Assessment for Changes to AI Agents
cs.AI, cs.CL, cs.LG
Submitted: 2026-10-12
Updated: 2026-10-12
Comments: 43 pages, 5 figures, 7 tables. Code, registrations, and logs: https://github.com/yzhao062/agent-risk
Code: https://github.com/yzhao062/agent-risk
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- $\tau^2$-Bench: Evaluating Conversational Agents in a Dual-Control Environment
- International AI Safety Report
- Agent Behavioral Contracts II: Certifying Compositional Reliability Without Assuming Independence
- Ctrl-Z: Controlling AI Agents via Resampling
- Chronicle: Cut-Point Replay for Regression Testing of LLM Agents
- Foundations of a Time-Consistent Counterfactual Actuarial Runtime for Autonomous AI Agents
- Security of AI Agents
- Cost Aware Best Arm Identification
- Measuring AI Ability to Complete Long Software Tasks
- Autoregressive Diffusion World Models for Off-Policy Evaluation of LLM Agents
- Large Language Models Often Know When They Are Being Evaluated
- What Can Be Enforced? A Theory of Certified Runtime Safety for Tool-Using Agents
- The AI risk repository: A meta-review, database, and taxonomy of risks from artificial intelligence
- Sequential stratified inference for the mean
- Adapting Probabilistic Risk Assessment for AI
- Crab: A Semantics-Aware Checkpoint/Restore Runtime for Agent Sandboxes
- When Agent Automation Becomes Profitable: Quantifying and Insuring Autonomous AI Risk through Trace-Economic Underwriting
- Qwen3 Technical Report
- A Survey on Trustworthy LLM Agents: Threats and Countermeasures
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