AI Agent Auditing

Ask about this topic

Overview

AI agent auditing studies whether autonomous systems follow their instructions and policies as they use tools, exchange information, and change external state. Researchers combine trace analysis, runtime monitoring and authorization, static and dynamic security testing, adversarial benchmarks, and tamper-evident logs to detect violations and reconstruct how they occurred. A central lesson is that correct final answers or benchmark scores do not establish safe or faithful execution; audits increasingly examine intermediate decisions, permissions, evidence provenance, and effects on the environment. The field still faces challenges in obtaining complete, independent records and reliable evaluators, while balancing stronger oversight with task utility, cost, and agent autonomy.

Latest papers

CardsList

269 papers79 papers in the last four weeks, up 39% on the four weeks before. 0.5% of all new papers.

  1. EnvOS: Auditing Computer-Use Agents When the World Moves On : Phenomenon-Based Task Construction, Certification, and Evidence-Based Evaluation

    Oct 29, 2026Ankush Kadu, Athira Gopal, Niharika Singh +2Computer-Use Agent EvaluationAI Agent Auditing

    QpiAI, Bangalore, India

  2. Agentic process evidence

    Oct 29, 2026Yuval Kainan, Carmit HershmanAI Agent GovernanceAI Agent Auditing

    JFrog

  3. Soei: Structured Observability of Evidence and Intent for Multi-Agent AI Software Engineering

    Oct 29, 2026Mohammed Nayeem Khan, K S Sarun KumarAI Agent MonitoringAI Coding Agents

    Intel Corporation

  4. Clean Sheets and Checkpoints: Evidence Complexity for Certifying Sequential Agent Risk

    Oct 29, 2026Rohith Yanambaka Venkata, Don McBride, Tiago Sequeira +1AI Agent AuditingAI Safety Evaluation

    Independent Researcher, United States · Cyber Defense Center, Nokia.

  5. Beyond fabrication: Misattributed numbers in papers written by AI research agents

    Oct 28, 2026Mosab Alfaqeeh, Omar Hourani, Muath Hourani +1Claim VerificationAI Agent Auditing

    Al-Ahliyya Amman University, Amman, Jordan · Binghamton University, State University of New York, Binghamton, NY, USA

  6. Tracekit Box: Running AI Agents in Micro-VMs Where Every Way Out Is Decided and Signed

    Oct 28, 2026Bravish GhoshRuntime Enforcement for AI AgentsAI Agent Security

    Cygnux Labs

  7. Detecting Covert Collusion between AI Agents through Partner-Outsider Positive Listening

    Oct 27, 2026Kevin Yandoka DenamganaïAI Agent MonitoringAI Agent Auditing

  8. Blindfold: Keeping the Audit Coin Secret from LLM Agents

    Oct 27, 2026Wei Zhu, Meiyun Zuo, Huiping SunAI Agent AuditingAI Agent Security

    School of Information, Renmin University of China, Beijing, 100872, China · Big Data and Responsible Artificial Intelligence for National Governance, Renmin University of China, Beijing, 100872, China · School of software and microelectronics, Peking University, Beijing, 100871, China

  9. Monitor Accuracy Isn't the Answer: A Decision-Value Framework for Evaluating AI Agent Restart Monitors

    Oct 27, 2026Siddharth Jain, Venkat N. VedamAI Agent MonitoringAI Agent Evaluation

    Intuit

  10. Identity or Evidence? Auditing Evidence Dependence in Agentic Life-Science Benchmarks

    Oct 27, 2026Danai Brilli, Maria GeorgilaScientific Agent BenchmarksEvidence-Grounded Reasoning

    Zo Labs

  11. FIDES-Live: Evaluating Behavioral Concordance in Sequential Trading Agents

    Oct 27, 2026Arther Tian, Alex Ding, Simon Wu +1AI Agent AuditingProcess-Based Agent Evaluation

    DGrid AI

  12. LLM WikiLeaks: Uncovering URL-Based Capability Extension and Security Circumvention in Agentic Message Boards

    Oct 25, 2026Orel Navi, Elad Feldman, Ben NassiLLM Agent SecurityWeb Agents

    Tel Aviv University, Tel Aviv, Israel

  13. PASS Is a Claim, Not an Observation: Mechanical Re-verification of 10 Self-Asserted Fix Claims in a Live Multi-Agent Audit Layer

    Oct 24, 2026Lei Zhenbing, Lei ChangAI Agent AuditingLLM Agent Verification

  14. Minimum Description Length Unsupervisedly Separates Law-Grade from Statistical-Grade Knowledge in a Self-Improving Science Agent - and Reveals 86% of Its Own Discoveries Are Statistical

    Oct 23, 2026Wenbin Mao, Pei ZouScientific DiscoveryAutonomous Scientific Discovery

    Artificial Intelligence Laboratory, Hooman Biotechnology (Beijing) Co., Ltd.

  15. Who Verifies the Tools? How Warped Formatters Mislead Agents and Their Verifiers

    Oct 23, 2026William JeffriesLLM Agent ReliabilityLLM Agent Verification

  16. Can an Agent Debug a Web Corpus? Agentic Auditing of Slovak FineWeb2

    Oct 22, 2026Marek Šuppa, Ivan Vykopal, Jaroslav KopčanAI Agent AuditingLanguage Model Auditing

    Comenius University in Bratislava · Cisco Systems · Kempelen Institute of Intelligent Technologies (KInIT) +1

  17. Contestable Agents for Data Governance: Recording Human Stewardship of AI-Generated Metadata Across Structured and Unstructured Systems

    Oct 22, 2026Humza Akhtar, Brian FreemanAI Agent GovernanceAI Governance

    Industry Solutions, MongoDB, Inc. · Advanced Concepts & Capabilities, Trane Technologies

  18. AfterDeleteBench: Auditing Post-Deletion Information Persistence in LLM Agents

    Oct 22, 2026Ehsan Saleh, Roozbeh Nahavandi, Sama Salarian +2LLM Agent EvaluationMemory Management for LLM Agents

    The Ohio State University

  19. AgentTriage: Replayable Security Diagnosis Across Heterogeneous Agent Runtimes

    Oct 22, 2026Jie Ying, Xutao Zhang, Ningyuan Chen +4AI Agent AuditingAI Agent Security

    College of Information and Intelligence Engineering, Zhejiang Wanli University, Ningbo, China · College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou, China

  20. Auditing the Product, Not the Recursion: The Evaluation Evidence Behind Recursive Self-Improvement Claims

    Oct 20, 2026Zihan Chen, Di Zhu, Lei Nico ZhengRecursive Self-ImprovementAI Agent Auditing

    Stevens Institute of Technology Hoboken, NJ, USA · University of Massachusetts Boston Boston, MA, USA

  21. LeakBench: Auditing Reward Hacking through Prediction-Time Leakage in Clinical Modeling Agents

    Oct 19, 2026Xiaoyuan Bao, Jingsi Jiang, Jiao Shan +4Data LeakageAI Agent Auditing

    Medical Informatics Center, Institute of Advanced Clinical Medicine, Peking University, Beijing, China · Department of Ultrasound, Peking University Third Hospital, Beijing, China · Department of Hospital-Acquired Infection Control, Beijing Jishuitan Hospital, Beijing, China +4

  22. A Cue's Hidden Price: Auditing Seller Influence on AI Shoppers

    Oct 19, 2026Felix Buecheler, Gerardo Berbeglia, Mingyu GuoAI Agent AuditingGenerative Engine Optimization

    Adelaide University, Adelaide, Australia · Melbourne Business School, Melbourne, Australia

  23. Auditing Tool Traces in Agentic LLMs: Source Localization and Counterfactual Dependence Testing

    Oct 18, 2026Junhao Liu, Tianxiao Li, Yongxian Wu +4LLM AuditingAI Agent Auditing

    University of California, Irvine · NEC Labs America

  24. Write Once, Check Many: Reusable Ledgers for Post-Hoc Policy Compliance Auditing of Agent Trajectories

    Oct 17, 2026Ashwin Nagarajan, Gursimran Panesar, Alicia Yang +3AI Agent AuditingLLM Agent Efficiency

    Carnegie Mellon University

  25. Shortcut Sensitivity in Agent-Safety Evaluations

    Oct 15, 2026Hasan Demirkiran, Jens ErnstbergerAI Safety EvaluationAI Agent Auditing

    Kontext

  26. PROBE: Learning to Audit Policy Compliance in Tool-Using LLM Agents

    Oct 15, 2026Kshitij Mishra, Abhijith Sharma, Nils Lukas +1AI Agent AuditingLLM Agent Evaluation

    Mohamed bin Zayed University of Artificial Intelligence, Abu Dhabi, UAE

  27. AgentEval: Evidence-Bounded Assessment of Stateful Agents

    Oct 15, 2026Ruiqi GaoAI Agent EvaluationAgent Evaluation

    Alibaba Cloud, Alibaba Group

  28. Approve What, Exactly? Effect Binding and Trusted Rendering for Agent Step-Up Authorization

    Oct 15, 2026Sai Varun Kodathala, Prashanth Pollishetty, Jaylen CargillAI Agent SecurityTool Access Control for LLM Agents

    SportsVision AI Minnetonka, MN

  29. BLADE: Auditing Information Use for Failure Attribution in LLM Multi-Agent Systems

    Oct 13, 2026Xiaoguang Guo, Yaohao Long, Chuxu ZhangFailure Attribution for LLM AgentsAgent Failure Analysis

    University of Connecticut

  30. AgentStripes: An Evidence-Graded Maturity Model for AI Agents

    Oct 13, 2026Nutan Kumar Naik, Vijay Prasad Poudel, Aditya Kumar Saroj +2AI Agent EvaluationAI Agent Auditing

    Algen.AI, Bengaluru, Karnataka, India · Department of Earth and Atmospheric Sciences, National Institute of Technology Rourkela, Odisha, India