Welcome to AEGIS 2026

The 1st International Workshop on
Agentic Edge and Generative Intelligent Systems

Co-located with IEEE MASS 2026 · Hong Kong SAR, China · October 21–23, 2026

Motivation

The rapid advancement of Large Language Models (LLMs), Generative AI, and Foundation Models is catalyzing a paradigm shift—from passive deep-learning inference toward autonomous Agentic AI. While traditional edge intelligence research has concentrated on model compression and training offloading, the next generation of mobile ad-hoc and smart systems demands intelligent agents capable of perception, reasoning, planning, and dynamic tool invocation directly at the network edge.

This evolution calls for a fundamental rethinking of edge computing architectures. We envision a future of Agentic Edge Intelligence, in which distributed edge devices—spanning robots, drones, smart vehicles, and IoT gateways—operate as autonomous agents. These agents must optimize local objectives using techniques such as reinforcement learning and diffusion models, while engaging in sophisticated coordination within Multi-Agent Systems (MAS). Meanwhile, breakthroughs in generative AI are unlocking entirely new capabilities at the edge, from content synthesis and data augmentation to generative planning for embodied intelligence.

How AEGIS 2026 Differs from the Main Conference

While IEEE MASS covers mobile ad-hoc and smart systems broadly, AEGIS 2026 specifically targets the emerging intersection of agentic AI, generative foundation models, and edge intelligence—a rapidly evolving area that falls between the boundaries of traditional systems, networking, and AI venues. The workshop provides a focused forum for topics such as LLM-based autonomous agents at the edge, diffusion model deployment on resource-constrained devices, and multi-agent collaborative intelligence, which are not specifically addressed by the main conference program.

Scope of Topics

We solicit original contributions in five major areas — see the Call for Papers for the full list:

• Foundations of Agentic Edge Intelligence (lightweight LLMs/SLMs, edge–cloud co-design, NPU/TPU optimization)
• Multi-Agent Systems and Collaborative Intelligence (LLM-based orchestration, game-theoretic allocation, agent-to-agent protocols)
• Generative AI and Diffusion Models at the Edge (edge-native content generation, generative policies for robotics, personalized generative agents)
• Edge Intelligence Optimization and Orchestration (DRL-based scheduling, energy-aware green AI, QoS for agentic interactions)
• Security, Privacy, and Trust in Agentic Systems (adversarial robustness, differential privacy, blockchain-based verification)

Target Communities

• Edge / Fog / Mobile Computing — IEEE MASS, SEC, MobiCom, MobiSys
• AI / ML Systems — MLSys, OSDI, NeurIPS Systems
• Multi-Agent Systems & Robotics — AAMAS, IROS, CoRL
• Networking & Communications — IEEE INFOCOM, JSAC, TMC
• Security & Privacy — adversarial robustness, trustworthy AI, distributed privacy
• Industry Practitioners — engineers deploying AI agents on edge platforms