pith. sign in

arxiv: 2606.12835 · v1 · pith:EG2O2S5Xnew · submitted 2026-06-11 · 💻 cs.MA · cs.AI· cs.CY· cs.NI

The Internet of Agentic AI: Communication, Coordination, and Collective Intelligence at Scale

Quanyan Zhu This is my paper

Pith reviewed 2026-06-27 05:29 UTC · model grok-4.3

classification 💻 cs.MA cs.AIcs.CYcs.NI
keywords Internet of Agentic AImulti-agent systemsautonomous agentscollective intelligenceagent coordinationsemantic interoperabilitysecure identitygovernance
0
0 comments X

The pith

Autonomous AI agents form an open ecosystem called the Internet of Agentic AI where they discover each other, negotiate responsibilities, exchange context, and execute workflows at scale.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops the vision of the Internet of Agentic AI as a distributed system of reasoning, communication, and action that extends beyond isolated models. It synthesizes foundations from multi-agent systems, distributed computing, game theory, and security engineering to outline deployment models, protocols, interoperability layers, and trust architectures. A sympathetic reader would care because this structure could enable collective intelligence across cloud, edge, device, and cyber-physical settings, as illustrated by case studies in adaptive manufacturing and operational coordination. The framework identifies six central research challenges that must be solved for such networks to function reliably.

Core claim

The paper characterizes the architectures and mechanisms required for scalable agent ecosystems by examining agent deployment models, workflow lifecycles, communication protocols, interoperability layers, resource-management challenges, and trust architectures, resulting in a framework that highlights the central research challenges of controlled emergence, semantic interoperability, secure identity, incentive-compatible coordination, resource-aware orchestration, and governance for large-scale networks of autonomous agents.

What carries the argument

The Internet of Agentic AI (IoAI) vision, which integrates single-agent agentic AI with multi-agent systems, networks, and security to enable heterogeneous agents to discover one another, negotiate, exchange context, invoke tools, and execute workflows across environments.

If this is right

  • Case studies show that agent workflows can adapt in manufacturing and distributed operational coordination when the listed mechanisms are present.
  • Controlled emergence and semantic interoperability become prerequisites for any large-scale agent network to avoid chaotic behavior.
  • Secure identity and incentive-compatible coordination determine whether agents will participate and maintain trust across organizational boundaries.
  • Resource-aware orchestration and governance structures are required to prevent overload and ensure accountability as the number of agents grows.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The IoAI model could naturally extend to agents that directly control physical actuators in cyber-physical systems, requiring additional safety layers not detailed in the review.
  • Standardized semantic protocols might need to evolve through iterative real-world trials rather than top-down design.
  • Economic incentives for agents could create new markets for coordination services similar to existing cloud marketplaces.
  • Governance challenges may intersect with existing regulatory frameworks for AI and data sharing in ways the synthesis does not yet address.

Load-bearing premise

Synthesizing existing foundations from multi-agent systems, networks, and security will sufficiently characterize the architectures needed for scalable agent ecosystems without first identifying specific technical barriers to interoperability at scale.

What would settle it

A concrete large-scale deployment of heterogeneous agents that exhibits uncontrolled emergence or fails semantic interoperability despite following the synthesized architectures would falsify the claim that existing foundations are adequate.

Figures

Figures reproduced from arXiv: 2606.12835 by Quanyan Zhu.

Figure 1
Figure 1. Figure 1: Conceptual architecture of an agentic AI system. The AI agent interacts with users, [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Conceptual illustration of IoAI. Autonomous AI agents communicate and coordinate [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Distributed agent deployment in IoAI. Specialized agents operate across cloud data cen [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Collaborative drug discovery in IoAI. A human researcher submits a scientific goal, a dis [PITH_FULL_IMAGE:figures/full_fig_p015_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: IoAI for manufacturing systems: decentralized coalition selection among heterogeneous [PITH_FULL_IMAGE:figures/full_fig_p040_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Distributed agentic coordination for Mosaic Warfare. Named space, air, and ground [PITH_FULL_IMAGE:figures/full_fig_p045_6.png] view at source ↗
read the original abstract

The rapid emergence of autonomous AI agents is transforming artificial intelligence from isolated model inference into distributed systems of reasoning, communication, and action. This paper develops the vision of the Internet of Agentic AI (IoAI): an open ecosystem in which heterogeneous agents discover one another, negotiate responsibilities, exchange context, invoke tools, and execute workflows across cloud, edge, device, organizational, and cyber-physical environments. We synthesize foundations from single-agent agentic AI, multi-agent systems, distributed computing, communication networks, game theory, and security engineering to characterize the architectures and mechanisms required for scalable agent ecosystems. The paper examines agent deployment models, workflow lifecycles, communication protocols, interoperability layers, resource-management challenges, and trust architectures, with case studies in adaptive manufacturing and distributed operational coordination. The resulting framework highlights the central research challenges of controlled emergence, semantic interoperability, secure identity, incentive-compatible coordination, resource-aware orchestration, and governance for large-scale networks of autonomous agents.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper develops a vision for the Internet of Agentic AI (IoAI) as an open ecosystem of heterogeneous autonomous agents that discover one another, negotiate, exchange context, and execute workflows across diverse environments. It synthesizes foundations from single-agent AI, multi-agent systems, distributed computing, networks, game theory, and security to characterize required architectures and mechanisms, examines deployment models, workflow lifecycles, protocols, interoperability layers, resource management, and trust architectures, includes case studies in adaptive manufacturing and distributed operational coordination, and identifies central research challenges of controlled emergence, semantic interoperability, secure identity, incentive-compatible coordination, resource-aware orchestration, and governance.

Significance. A coherent high-level framework that accurately maps existing concepts onto scalable agent ecosystems and correctly prioritizes open problems could help organize research in multi-agent systems and distributed AI. Its value would lie in synthesis and problem identification rather than new mechanisms, proofs, or empirical results.

major comments (2)
  1. [Abstract] Abstract: the claim that the synthesis from the listed fields 'characterizes the architectures and mechanisms required for scalable agent ecosystems' is not supported by any indication of new gap analysis, derivation of concrete requirements, or identification of specific scalability barriers (e.g., quantitative bounds on coordination overhead or semantic interoperability failure modes); the text instead enumerates existing concepts, leaving the highlighted challenges as assertions rather than derived conclusions.
  2. [Abstract] Abstract and case-study description: the referenced case studies in adaptive manufacturing and distributed operational coordination are presented as illustrations, yet the abstract provides no evidence that they include concrete metrics, protocol evaluations at scale, or tests of when existing mechanisms break, which would be needed to ground the listed research challenges.
minor comments (1)
  1. [Abstract] The term 'agentic AI' is used without an explicit definition or citation to its originating literature in the abstract; a brief clarification would improve accessibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We agree that the manuscript is a vision and synthesis paper whose primary contributions are mapping existing concepts across fields and identifying open challenges, rather than deriving new quantitative analyses or empirical evaluations. We will revise the abstract accordingly to ensure the language accurately reflects this scope.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that the synthesis from the listed fields 'characterizes the architectures and mechanisms required for scalable agent ecosystems' is not supported by any indication of new gap analysis, derivation of concrete requirements, or identification of specific scalability barriers (e.g., quantitative bounds on coordination overhead or semantic interoperability failure modes); the text instead enumerates existing concepts, leaving the highlighted challenges as assertions rather than derived conclusions.

    Authors: We accept this point. The abstract's wording implies a stronger derivation of requirements than the synthesis provides. The manuscript maps concepts from the listed fields to surface challenges but does not include new gap analyses or quantitative bounds. We will revise the abstract to state that the synthesis 'outlines' architectures and mechanisms and 'identifies' challenges through this mapping, removing any implication of novel derivation or concrete scalability barriers. revision: yes

  2. Referee: [Abstract] Abstract and case-study description: the referenced case studies in adaptive manufacturing and distributed operational coordination are presented as illustrations, yet the abstract provides no evidence that they include concrete metrics, protocol evaluations at scale, or tests of when existing mechanisms break, which would be needed to ground the listed research challenges.

    Authors: We agree. The case studies function as conceptual illustrations to motivate the challenges and are not accompanied by metrics, large-scale evaluations, or failure-mode tests. We will revise the abstract to explicitly describe them as 'illustrative case studies' and clarify that the research challenges are identified via the overall framework rather than empirically grounded by the examples. revision: yes

Circularity Check

0 steps flagged

No circularity: high-level synthesis paper without derivations or predictions

full rationale

The paper is a vision and synthesis document that enumerates architectures, challenges, and case studies by drawing on existing literature from MAS, networks, game theory, and security. It contains no equations, no fitted parameters, no predictions of quantities, and no load-bearing uniqueness theorems or ansatzes. The central activity is high-level characterization via synthesis, which does not reduce any claimed result to its own inputs by construction. Self-citations, if present, are not used to justify a derivation that would otherwise be circular. This is the normal, non-circular outcome for a survey-style framework paper.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

As a vision paper, the central claim rests on domain assumptions from multi-agent systems and distributed computing without introducing new free parameters, axioms, or invented entities in the abstract.

pith-pipeline@v0.9.1-grok · 5697 in / 996 out tokens · 21074 ms · 2026-06-27T05:29:58.054396+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

68 extracted references · 25 canonical work pages · 9 internal anchors

  1. [1]

    Agent2Agent protocol specification.https://a2a-protocol.org/latest/spe cification/, 2026

    A2A Project. Agent2Agent protocol specification.https://a2a-protocol.org/latest/spe cification/, 2026. Latest specification

    2026

  2. [2]

    Abou Ali, F

    M. Abou Ali, F. Dornaika, and J. Charafeddine. Agentic AI: A comprehensive survey of architectures, applications, and future directions.Artificial Intelligence Review, 59:11, 2026

    2026

  3. [3]

    Concrete Problems in AI Safety

    D. Amodei, C. Olah, J. Steinhardt, P. Christiano, J. Schulman, and D. Man´ e. Concrete problems in AI safety.arXiv preprint arXiv:1606.06565, 2016

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  4. [4]

    Asgar, M

    Z. Asgar, M. Nguyen, and S. Katti. Efficient and scalable agentic AI with heterogeneous systems.arXiv preprint arXiv:2507.19635, 2025

    work page arXiv 2025

  5. [5]

    Y. Bai, S. Kadavath, S. Kundu, A. Askell, J. Kernion, A. Jones, A. Chen, A. Goldie, A. Mirho- seini, C. McKinnon, C. Chen, C. Olsson, C. Olah, D. Hernandez, D. Drain, D. Ganguli, D. Li, E. Tran-Johnson, E. Perez, J. Kerr, J. Mueller, J. Ladish, J. Landau, K. Ndousse, K. Lukosuite, L. Lovitt, M. Sellitto, N. Elhage, N. Schiefer, N. Mercado, N. DasSarma, R....

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  6. [6]

    Bandi, B

    A. Bandi, B. Kongari, R. Naguru, S. Pasnoor, and S. V. Vilipala. The rise of agentic AI: A re- view of definitions, frameworks, architectures, applications, evaluation metrics, and challenges. Future Internet, 17(9):404, 2025

    2025

  7. [7]

    Ba¸ sar and G

    T. Ba¸ sar and G. J. Olsder.Dynamic Noncooperative Game Theory. SIAM, 2nd edition, 1999

    1999

  8. [8]

    J. Benet. IPFS – content addressed, versioned, P2P file system.arXiv preprint arXiv:1407.3561, 2014

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  9. [9]

    Biggio and F

    B. Biggio and F. Roli. Wild patterns: Ten years after the rise of adversarial machine learning. In Proceedings of the 2018 ACM SIGSAC Conference on Computer and Communications Security, pages 2154–2156. ACM, 2018

    2018

  10. [10]

    Bonabeau, M

    E. Bonabeau, M. Dorigo, and G. Theraulaz.Swarm Intelligence: From Natural to Artificial Systems. Oxford University Press, 1999

    1999

  11. [11]

    Bourgoyne

    A. Bourgoyne. Scaling autonomous AI agents and workloads with NVIDIA DGX spark.https: //developer.nvidia.com/blog/scaling-autonomous-ai-agents-and-workloads-with-n vidia-dgx-spark, 2026. NVIDIA technical blog

    2026

  12. [12]

    V. Buterin. Ethereum: A next-generation smart contract and decentralized application plat- form.https://ethereum.org/en/whitepaper/, 2014. White paper

    2014

  13. [13]

    Camazine, J.-L

    S. Camazine, J.-L. Deneubourg, N. Franks, J. Sneyd, G. Theraulaz, and E. Bonabeau.Self- Organization in Biological Systems. Princeton University Press, 2003

    2003

  14. [14]

    Chalkiadakis, E

    G. Chalkiadakis, E. Elkind, and M. Wooldridge.Computational Aspects of Cooperative Game Theory. Morgan & Claypool, 2011

    2011

  15. [15]

    Chang, E

    G. Chang, E. Lin, C. Yuan, R. Cai, B. Chen, X. Xie, and Y. Zhang. Agent network protocol technical white paper.arXiv preprint arXiv:2508.00007, 2025

    work page arXiv 2025

  16. [16]

    Chen and Q

    J. Chen and Q. Zhu. A games-in-games approach to mosaic command and control design of dynamic network-of-networks for secure and resilient multi-domain operations.arXiv preprint arXiv:1906.02114, 2019

    work page arXiv 1906

  17. [17]

    Z. S. Chen and Q. Zhu. A theory of multilevel interactive equilibrium in NeuroAI.arXiv preprint arXiv:2605.10505, 2026

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  18. [18]

    Costan and S

    V. Costan and S. Devadas. Intel SGX explained.IACR Cryptology ePrint Archive, 2016:86, 2016

    2016

  19. [19]

    MATHBAC: Mathematics of boosting agentic communication.https://www.darpa.mil/research/programs/mathbac, 2026

    Defense Advanced Research Projects Agency. MATHBAC: Mathematics of boosting agentic communication.https://www.darpa.mil/research/programs/mathbac, 2026. Program description

    2026

  20. [20]

    J. R. Douceur. The sybil attack. InProceedings of the 1st International Workshop on Peer-to- Peer Systems, volume 2429 ofLecture Notes in Computer Science, pages 251–260. Springer, 2002

    2002

  21. [21]

    Duan and Z

    Q. Duan and Z. Lu. Agent communications toward agentic AI at edge – a case study of the Agent2Agent protocol.arXiv preprint arXiv:2508.15819, 2025

    work page arXiv 2025

  22. [22]

    P. T. Eugster, P. A. Felber, R. Guerraoui, and A.-M. Kermarrec. The many faces of publish/- subscribe.ACM Computing Surveys, 35(2):114–131, 2003

    2003

  23. [23]

    Giusti, O

    L. Giusti, O. A. Werner, R. Taiello, M. C. Costa, E. Tosun, A. Protani, M. Molina, R. Lopes de Almeida, P. Cacace, D. R. Santos, and L. Serio. Federation of agents: A semantics-aware communication fabric for large-scale agentic AI.arXiv preprint arXiv:2509.20175, 2025

    work page arXiv 2025

  24. [24]

    Han and Q

    Z. Han and Q. Zhu. Toward a dynamic stackelberg game-theoretic framework for agentic AI defense against LLM jailbreaking.arXiv preprint arXiv:2507.08207, 2025. Accepted to ICLR 2026 AIMS Workshop

    work page arXiv 2025

  25. [25]

    J. H. Holland.Complexity: A Very Short Introduction. Oxford University Press, 2014

    2014

  26. [26]

    S. Hong, M. Zhuge, J. Chen, X. Zheng, Y. Cheng, C. Zhang, J. Wang, Z. Wang, S. K. S. Yau, Z. Lin, L. Zhou, C. Ran, L. Xiao, C. Wu, and J. Schmidhuber. MetaGPT: Meta programming for a multi-agent collaborative framework.arXiv preprint arXiv:2308.00352, 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  27. [27]

    Huang, V

    K. Huang, V. S. Narajala, J. Yeoh, J. Ross, M. Lambe, R. Raskar, Y. Harkati, J. Huang, I. Habler, and C. Hughes. A novel zero-trust identity framework for agentic AI: Decentralized authentication and fine-grained access control.arXiv preprint arXiv:2505.19301, 2025

    work page arXiv 2025

  28. [28]

    Agentic AI for next generation wireless networks.https: //www.comsoc.org/publications/magazines/ieee-network/cfp/agentic-ai-next-gen eration-wireless-networks, 2024

    IEEE Communications Society. Agentic AI for next generation wireless networks.https: //www.comsoc.org/publications/magazines/ieee-network/cfp/agentic-ai-next-gen eration-wireless-networks, 2024. IEEE Network special issue call for papers

    2024

  29. [29]

    Kairouz, H

    P. Kairouz, H. B. McMahan, B. Avent, A. Bellet, M. Bennis, A. N. Bhagoji, K. Bonawitz, Z. Charles, G. Cormode, R. Cummings, R. G. L. D’Oliveira, H. Eichner, S. El Rouayheb, D. Evans, J. Gardner, Z. Garrett, A. Gasc´ on, B. Ghazi, P. B. Gibbons, M. Gruteser, Z. Har- chaoui, C. He, L. He, Z. Huo, B. Hutchinson, J. Hsu, M. Jaggi, T. Javidi, G. Joshi, M. Khod...

    2021

  30. [30]

    Kamatar, J

    A. Kamatar, J. G. Pauloski, Y. Babuji, R. Chard, M. Sakarvadia, D. Babnigg, K. Chard, and I. Foster. Empowering scientific workflows with federated agents.arXiv preprint arXiv:2505.05428, 2025

    work page arXiv 2025

  31. [31]

    3 things to know about keyfactor’s PKI-based identity for agentic AI.https: //www.keyfactor.com/blog/3-things-to-know-about-keyfactors-pki-based-identit y-for-agentic-ai, 2025

    Keyfactor Team. 3 things to know about keyfactor’s PKI-based identity for agentic AI.https: //www.keyfactor.com/blog/3-things-to-know-about-keyfactors-pki-based-identit y-for-agentic-ai, 2025. Technical blog

    2025

  32. [32]

    N. K. Krishnan. Beyond context sharing: A unified agent communication protocol (ACP) for secure, federated, and autonomous agent-to-agent (A2A) orchestration.arXiv preprint arXiv:2602.15055, 2026

    work page arXiv 2026

  33. [33]

    L. Lamport. The part-time parliament.ACM Transactions on Computer Systems, 16(2):133– 169, 1998

    1998

  34. [34]

    LangGraph overview.https://docs.langchain.com/oss/python/langgraph/ overview, 2026

    LangChain. LangGraph overview.https://docs.langchain.com/oss/python/langgraph/ overview, 2026. Documentation

    2026

  35. [35]

    G. Li, H. A. A. K. Hammoud, H. Itani, D. Khizbullin, and B. Ghanem. CAMEL: Com- municative agents for ”mind” exploration of large language model society.arXiv preprint arXiv:2303.17760, 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  36. [36]

    Li and Q

    T. Li and Q. Zhu. Agentic AI for cyber resilience: A new security paradigm and its system- theoretic foundations.arXiv preprint arXiv:2512.22883, 2025

    work page arXiv 2025

  37. [37]

    X. Liu, J. Zhang, H. Shang, S. Guo, C. Yang, and Q. Zhu. Exploring prosocial irrationality for LLM agents: A social cognition view.arXiv preprint arXiv:2405.14744, 2025. Published as a conference paper at ICLR 2025

    work page arXiv 2025

  38. [38]

    Marro, E

    S. Marro, E. La Malfa, J. Wright, G. Li, N. Shadbolt, M. Wooldridge, and P. Torr. A scalable communication protocol for networks of large language models.arXiv preprint arXiv:2410.11905, 2024

    work page arXiv 2024

  39. [39]

    Agent framework overview.https://learn.microsoft.com/en-us/agent-frame work/overview/, 2026

    Microsoft. Agent framework overview.https://learn.microsoft.com/en-us/agent-frame work/overview/, 2026. Microsoft documentation

    2026

  40. [40]

    MITRE ATLAS: Adversarial threat landscape for artificial-intelligence systems, 2025

    MITRE. MITRE ATLAS: Adversarial threat landscape for artificial-intelligence systems, 2025. Accessed 2026

    2025

  41. [41]

    Model context protocol specification.https://modelcontextprot ocol.io/specification/2025-11-25, 2025

    Model Context Protocol. Model context protocol specification.https://modelcontextprot ocol.io/specification/2025-11-25, 2025. Version 2025-11-25

    2025

  42. [42]

    Security and privacy controls for information systems and organizations

    National Institute of Standards and Technology. Security and privacy controls for information systems and organizations. NIST Special Publication 800-53 Revision 5, 2020. Updated 2023

    2020

  43. [43]

    Zero trust architecture

    National Institute of Standards and Technology. Zero trust architecture. NIST Special Publi- cation 800-207, 2020

    2020

  44. [44]

    Artificial intelligence risk management frame- work: Generative artificial intelligence profile

    National Institute of Standards and Technology. Artificial intelligence risk management frame- work: Generative artificial intelligence profile. NIST AI 600-1, 2024

    2024

  45. [45]

    Olfati-Saber, J

    R. Olfati-Saber, J. A. Fax, and R. M. Murray. Consensus and cooperation in networked multi-agent systems.Proceedings of the IEEE, 95(1):215–233, 2007

    2007

  46. [46]

    Ongaro and J

    D. Ongaro and J. Ousterhout. In search of an understandable consensus algorithm. InPro- ceedings of the 2014 USENIX Annual Technical Conference, pages 305–319, 2014

    2014

  47. [47]

    Introducing ChatGPT agent: Bridging research and action.https://openai.com /index/introducing-chatgpt-agent/, 2025

    OpenAI. Introducing ChatGPT agent: Bridging research and action.https://openai.com /index/introducing-chatgpt-agent/, 2025. Product announcement

    2025

  48. [48]

    New tools for building agents.https://openai.com/index/new-tools-for-build ing-agents/, 2025

    OpenAI. New tools for building agents.https://openai.com/index/new-tools-for-build ing-agents/, 2025. Product announcement

    2025

  49. [49]

    OWASP top 10 for large language model applications 2025, 2025

    OWASP Foundation. OWASP top 10 for large language model applications 2025, 2025

    2025

  50. [50]

    Panait and S

    L. Panait and S. Luke. Cooperative multi-agent learning: The state of the art.Autonomous Agents and Multi-Agent Systems, 11(3):387–434, 2005

    2005

  51. [51]

    J. S. Park, J. O’Brien, C. J. Cai, M. R. Morris, P. Liang, and M. S. Bernstein. Genera- tive agents: Interactive simulacra of human behavior.Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology, 2023

    2023

  52. [52]

    Ren and R

    W. Ren and R. W. Beard.Distributed Consensus in Multi-Vehicle Cooperative Control. Springer, 2007

    2007

  53. [53]

    AI Agents with Decentralized Identifiers and Verifiable Credentials

    S. Rodriguez Garzon, A. Vaziry, E. M. Kuzu, D. E. Gehrmann, B. Varkan, A. Gaballa, and A. K¨ upper. AI agents with decentralized identifiers and verifiable credentials.arXiv preprint arXiv:2511.02841, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  54. [54]

    Russell and P

    S. Russell and P. Norvig.Artificial Intelligence: A Modern Approach. Pearson, 4th edition, 2021

    2021

  55. [55]

    Open Challenges in Multi-Agent Security: Towards Secure Systems of Interacting AI Agents

    C. Schroeder de Witt. Open challenges in multi-agent security: Towards secure systems of interacting AI agents.arXiv preprint arXiv:2505.02077, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  56. [56]

    Shapiro, N

    M. Shapiro, N. Pregui¸ ca, C. Baquero, and M. Zawirski. Conflict-free replicated data types. InProceedings of the 13th International Symposium on Stabilization, Safety, and Security of Distributed Systems, pages 386–400, 2011

    2011

  57. [57]

    Shoham and K

    Y. Shoham and K. Leyton-Brown.Multiagent Systems: Algorithmic, Game-Theoretic, and Logical Foundations. Cambridge University Press, 2008

    2008

  58. [58]

    M. Silwer. Trust standards evolve: AI agents, the next chapter for PKI.https://blog.hid global.com/trust-standards-evolve-ai-agents-next-chapter-pki, 2025. HID Global blog

    2025

  59. [59]

    Stone and M

    P. Stone and M. Veloso. Multiagent systems: A survey from a machine learning perspective. Autonomous Robots, 8(3):345–383, 2000

    2000

  60. [60]

    Varghese and R

    B. Varghese and R. Buyya. Next generation cloud computing: New trends and research directions.Future Generation Computer Systems, 79:849–861, 2018

    2018

  61. [61]

    Wooldridge.An Introduction to MultiAgent Systems

    M. Wooldridge.An Introduction to MultiAgent Systems. Wiley, 2nd edition, 2009

    2009

  62. [62]

    Decentralized identifiers (DIDs) v1.0: Core architecture, data model, and representations.https://www.w3.org/TR/did-core/, 2022

    World Wide Web Consortium. Decentralized identifiers (DIDs) v1.0: Core architecture, data model, and representations.https://www.w3.org/TR/did-core/, 2022. W3C Recommenda- tion, 19 July 2022

    2022

  63. [63]

    Verifiable credentials data model v2.0.https://www.w3.org /TR/vc-data-model-2.0/, 2025

    World Wide Web Consortium. Verifiable credentials data model v2.0.https://www.w3.org /TR/vc-data-model-2.0/, 2025. W3C Recommendation, 15 May 2025

    2025

  64. [64]

    Yang and Q

    Y.-T. Yang and Q. Zhu. Differential privacy in generative AI agents: Analysis and optimal tradeoffs.arXiv preprint arXiv:2603.17902, 2026

    work page arXiv 2026

  65. [65]

    Yang and Q

    Y.-T. Yang and Q. Zhu. Internet of agentic AI: Incentive-compatible distributed teaming and workflow.arXiv preprint arXiv:2602.03145, 2026

    work page arXiv 2026

  66. [66]

    Q. Zhu. Game theory meets LLM and agentic AI: Reimagining cybersecurity for the age of intelligent threats.arXiv preprint arXiv:2507.10621, 2025

    work page arXiv 2025

  67. [67]

    Q. Zhu. LLM-stackelberg games: Conjectural reasoning equilibria and their applications to spearphishing.arXiv preprint arXiv:2507.09407, 2025

    work page arXiv 2025

  68. [68]

    Q. Zhu. Insurance of agentic AI.arXiv preprint arXiv:2606.05449, 2026. Table 4: Threat taxonomy for IoAI ecosystems. Threat Category Threat Type Description Affected Security Property Identity ThreatsSybil attacks An adversary creates multiple fake agent identities to gain disproportionate influence or reputation. Authenticity, accountability, trust [20,4...

    work page internal anchor Pith review Pith/arXiv arXiv 2026