Securing AI-Agentic Interactions via Multi-Agent Reinforcement Learning (MARL) with Secure Communication Protocols

Mohan Vamsi Musunuru () and Srinivas Bangalore Sujayendra Rao ()

Journal of Artificial Intelligence General science (JAIGS) ISSN:3006-4023, 2024, vol. 4, issue 1, 503-524

Abstract: The rapid deployment of AI-driven autonomous agents in critical applications has increased the importance of secure and reliable inter-agent communication. This study presents a novel framework that integrates Multi-Agent Reinforcement Learning (MARL) with secure communication protocols to enhance trust, privacy, and resilience in AI-agentic interactions. The proposed approach leverages decentralized training and policy sharing, while employing cryptographic techniques—such as end-to-end encryption and key exchange mechanisms—to safeguard information flow among agents. Experimental results in simulated environments demonstrate that our method not only maintains competitive task performance but also significantly reduces vulnerabilities to eavesdropping, message tampering, and adversarial manipulation. By aligning MARL coordination strategies with robust security mechanisms, this research provides a scalable and adaptive foundation for deploying multi-agent systems in high-stakes domains, including autonomous robotics, distributed IoT networks, and secure collaborative decision-making platforms.

Keywords: Multi-Agent Reinforcement Learning; AI-Agent Security; Secure Communication Protocols; Cryptographic Techniques; Privacy-Preserving AI; Distributed AI Systems; Agent Coordination; Adversarial Robustness (search for similar items in EconPapers)
Date: 2024
References: Add references at CitEc
Citations:

Downloads: (external link)
https://newjaigs.com/index.php/JAIGS/article/view/397 (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:das:njaigs:v:4:y:2024:i:1:p:503-524:id:397

Access Statistics for this article

Journal of Artificial Intelligence General science (JAIGS) ISSN:3006-4023 is currently edited by Justyna Żywiołek

More articles in Journal of Artificial Intelligence General science (JAIGS) ISSN:3006-4023 from Open Knowledge
Bibliographic data for series maintained by Open Knowledge ().