mlsanityData to 7 Oct 2026

Inherent and emergent liability issues in LLM-based agentic systems: a principal-agent perspective

arXiv 2504.03255 ↗First posted 4 Apr 2025cs.CY cs.CL cs.MAcited by 12

On the map: LLM agents and tool use → Agent threats, guardrails and governance

Abstract (arXiv)

Agentic systems powered by large language models (LLMs) are becoming progressively more complex and capable. Their increasing agency and expanding deployment settings attract growing attention to effective governance policies, monitoring, and control protocols. Based on the emerging landscape of the agentic market, we analyze potential liability issues arising from the delegated use of LLM agents and their extended systems through a principal-agent perspective. Our analysis complements existing risk-based studies on artificial agency and covers the spectrum of important aspects of the principal-agent relationship and their potential consequences at deployment. Furthermore, we motivate method developments for technical governance along the directions of interpretability and behavior evaluations, reward and conflict management, and the mitigation of misalignment and misconduct through principled engineering of detection and fail-safe mechanisms. By illustrating the outstanding issues in AI liability for LLM-based agentic systems, we aim to inform the system design, auditing, and tracing to enhance transparency and liability attribution.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Placed in Agent threats, guardrails and governance, one of 16 directions in LLM agents and tool use. Builds on 79 of its 170 references that are in the map. First posted on arXiv 4 Apr 2025.

Builds on

79 of 170 references.

Closest ancestors · 4

Same direction, newest first

  1. 2025Red-Teaming LLM Multi-Agent Systems via Communication Attacks20 Feb
  2. 2024GuardAgent: Safeguard LLM Agents by a Guard Agent via Knowledge-Enabled Reasoning13 Jun
  3. 2024Security of AI Agents12 Jun
  4. 2024TrustAgent: Towards Safe and Trustworthy LLM-based Agents2 Feb

Other references · 75

  1. 2019Machine Unlearningcited by 1,020
  2. 2023Steering Llama 2 via Contrastive Activation Additioncited by 1,009
  3. 2023Towards Understanding Sycophancy in Language Modelscited by 988
  4. 2022Discovering Language Model Behaviors with Model-Written Evaluationscited by 805
  5. 2024Evaluating Very Long-Term Conversational Memory of LLM Agentscited by 800
  6. 2024Large Language Model based Multi-Agents: A Survey of Progress and Challengescited by 752
  7. 2024Executable Code Actions Elicit Better LLM Agentscited by 537
  8. 2025Multi-Agent Collaboration Mechanisms: A Survey of LLMscited by 424
  9. 2023Cognitive Architectures for Language Agentscited by 422
  10. 2024Alignment faking in large language modelscited by 316
  11. 2023MedAgents: Large Language Models as Collaborators for Zero-shot Medical Reasoningcited by 311
  12. 2023Role-Play with Large Language Modelscited by 290
38 more
  1. 2023Faithful Chain-of-Thought Reasoningcited by 285
  2. 2024Knowledge Conflicts for LLMs: A Surveycited by 274
  3. 2024Rethinking Machine Unlearning for Large Language Modelscited by 247
  4. 2024Scaling Large Language Model-based Multi-Agent Collaborationcited by 235
  5. 2024LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individualscited by 235
  6. 2024From Persona to Personalization: A Survey on Role-Playing Language Agentscited by 206
  7. 2023Causal Abstraction: A Theoretical Foundation for Mechanistic Interpretabilitycited by 194
  8. 2024Foundational Challenges in Assuring Alignment and Safety of Large Language Modelscited by 193
  9. 2023AI Deception: A Survey of Examples, Risks, and Potential Solutionscited by 185
  10. 2024TradingAgents: Multi-Agents LLM Financial Trading Frameworkcited by 182
  11. 2025AgentSociety: Large-Scale Simulation of LLM-Driven Generative Agents Advances Understanding of Human Behaviors and Societycited by 180
  12. 2023Sociotechnical Safety Evaluation of Generative AI Systemscited by 154
  13. 2025Multi-Agent Risks from Advanced AIcited by 150
  14. 2024ProSA: Assessing and Understanding the Prompt Sensitivity of LLMscited by 145
  15. 2023Harms from Increasingly Agentic Algorithmic Systemscited by 134
  16. 2023Large Language Models can Strategically Deceive their Users when Put Under Pressurecited by 132
  17. 2024The Butterfly Effect of Altering Prompts: How Small Changes and Jailbreaks Affect Large Language Model Performancecited by 97
  18. 2024Visibility into AI Agentscited by 77
  19. 2024A Review of Prominent Paradigms for LLM-Based Agents: Tool Use (Including RAG), Planning, and Feedback Learningcited by 74
  20. 2023Turning large language models into cognitive modelscited by 66
  21. 2025Detecting Strategic Deception Using Linear Probescited by 65
  22. 2023Black Box Adversarial Prompting for Foundation Modelscited by 62
  23. 2018Incomplete Contracting and AI Alignmentcited by 60
  24. 2024Algorithmic Collusion by Large Language Modelscited by 54
  25. 2021Agent Incentives: A Causal Perspectivecited by 51
  26. 2024On Targeted Manipulation and Deception when Optimizing LLMs for User Feedbackcited by 45
  27. 2023Characterizing Manipulation from AI Systemscited by 43
  28. 2025Fully Autonomous AI Agents Should Not be Developedcited by 43
  29. 2022A Survey of Ad Hoc Teamwork Researchcited by 41
  30. 2025Infrastructure for AI Agentscited by 40
  31. 2019Ask Not What AI Can Do, But What AI Should Do: Towards a Framework of Task Delegabilitycited by 36
  32. 2020AI Research Considerations for Human Existential Safety (ARCHES)cited by 33
  33. 2023Exploring the Sensitivity of LLMs' Decision-Making Capabilities: Insights from Prompt Variation and Hyperparameterscited by 32
  34. 2024Towards evaluations-based safety cases for AI schemingcited by 32
  35. 2023Artificial Influence: An Analysis Of AI-Driven Persuasioncited by 31
  36. 2024On the Quest for Effectiveness in Human Oversight: Interdisciplinary Perspectivescited by 28
  37. 2024How Interpretable are Reasoning Explanations from Prompting Large Language Models?cited by 27
  38. 2025Multi-Agent Security Tax: Trading Off Security and Collaboration Capabilities in Multi-Agent Systemscited by 27

50 most cited of 75 All 75 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 7; 2026: 5.

12 papers.

  1. 2025AI Agents vs. Agentic AI: A Conceptual Taxonomy, Applications and Challengescited by 167
  2. 2025Agentic AI: A Comprehensive Survey of Architectures, Applications, and Future Directionscited by 46
  3. 2025A Call for Collaborative Intelligence: Why Human-Agent Systems Should Precede AI Autonomycited by 9
  4. 2025Towards Trustworthy Legal AI through LLM Agents and Formal Reasoningcited by 4
  5. 2025Reliable agent engineering should integrate machine-compatible organizational principlescited by 2
  6. 2025EU-Agent-Bench: Measuring Illegal Behavior of LLM Agents Under EU Lawcited by 1
  7. 2026Blockchain Empowered Trustworthy Agent Networks: Foundations, Taxonomy, and Future Directionscited by 1
  8. 2025Medication counseling with large language models: balancing flexibility and rigiditycited by 0
  9. 2026Distributed Legal Infrastructure for a Trustworthy Agentic Webcited by 0
  10. 2026An Agentic Framework Using Rules and LLMs for Embedding and Annotating Descriptive Document Layouts: A Plant Science Use Casecited by 0
  11. 2026The Civilization Framework: Sovereign-Anchored Communication Between Personal Multi-Agent Systemscited by 0
  12. 2026Et Tu, Brute? Economic Misalignment in Personal AI Agentscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Regulating the Agency of LLM-based Agents
  2. 2025AGENTSAFE: A Unified Framework for Ethical Assurance and Governance in Agentic AI
  3. 2024Practical Considerations for Agentic LLM Systems
  4. 2025Agentic AI Security: Threats, Defenses, Evaluation, and Open Challenges
  5. 2025Risk Analysis Techniques for Governed LLM-based Multi-Agent Systems
  6. 2025Interpretable Risk Mitigation in LLM Agent Systems
  7. 2025TRiSM for Agentic AI: A Review of Trust, Risk, and Security Management in LLM-based Agentic Multi-Agent Systems
  8. 2026LLM-Based Agents for Software and Systems Security: Approaches, Applications, and Assessment