Embedded Files
Workshop on
May 26, 2026
Overview
Critical National Infrastructure (CNI) forms the backbone of modern society, encompassing the vital systems required for the delivery of public services from energy grids and transportation networks to telecommunications and healthcare. While the reliability of these systems is a priority for both government and industry, their increasing complexity makes them susceptible to disruptions from technical failures, natural disasters, and human error.
The rapid evolution of Artificial Intelligence offers a transformative opportunity to enhance the operability of CNI. AI-driven solutions are being deployed to optimize resource distribution, predict maintenance needs before failures occur, and provide real-time responses to sophisticated cyber-physical threats. These advancements promise a future of more efficient, self-healing infrastructure that can adapt to the needs of a growing population.
However, integrating AI into these high-stakes environments is not a trivial task. As these systems move toward higher levels of autonomy, the role of AI Safety becomes a critical enabler of their deployment. To fully realize the benefits of AI, we must address the technical challenges of ensuring these systems remain robust, predictable, and transparent. This workshop seeks to explore the intersection of AI innovation and infrastructure resilience, focusing on:
Operational Excellence: Leveraging AI to increase the efficiency and reliability of vital service delivery.
Defense & Mitigation: Using AI to repel cyberattacks and minimize the impact of natural disturbances on the network.
AI Safety & Assurance: Developing frameworks to ensure that increasingly autonomous systems operate safely under all conditions, maintaining public trust in the national infrastructure.
This workshop offers a natural connection to the AAMAS community, providing an opportunity to apply multi-agent research to the unique challenges of coordination, AI safety, and the high performance standards inherent to Critical National Infrastructure. Please use the main conference website to register: https://cyprusconferences.org/aamas2026/
Topics
By bringing together experts from different stakeholders, this workshop aims to define how we can safely harness the power of AI to build more resilient Critical National Infrastructure. Our goal is to provide a highly interactive session where attendees can present their latest research but also lead demonstrations or drive a discussion round. All accepted papers will be invited to present a poster and we aim to have most papers with an oral presentation (depending on number of submissions).
We welcome submissions on the following topics:
1. Sector-Specific AI and Multi-Agent Applications
Energy & Utilities: Smart grid optimization, distributed agent coordination for energy transmission resilience, and nuclear energy generation/nonproliferation.
Mobility & Logistics: Multi-agent fleet management for autonomous transportation, supply chain optimization, and waste management.
Public Health & Safety: AI for hospital operations, coordinated emergency response services, and water treatment/distribution systems.
Industrial Systems: AI-driven safety and human-agent collaboration in the chemical industry and heavy manufacturing.
2. Resilience, Defense, and Cybersecurity
Autonomous Defense: Multi-agent systems for the cybersecurity of critical systems and collective threat detection.
Incident Management: AI for the rapid detection, isolation, and automated agent-led repair of service interruptions.
Mitigation: Leveraging AI to minimize the impact of natural disturbances and preventing cascading failures across interconnected CNI sectors.
3. AI Safety, Reliability, and Trust
Safety-Critical Autonomy: Frameworks for ensuring the safety and predictability of autonomous agents and collectives in CNI.
System Reliability: Methods for formal verification, validation, and uncertainty quantification in multi-agent high-stakes environments.
Human-AI Interaction: Agents in government and public policy to ensure transparent and accountable decision-making.
4. Foundation and Methodology
Cross-Domain Research: Fundamental AI/ML research addressing the unique constraints of real-world CNI (e.g., low-latency requirements, data scarcity).
Simulations & Benchmarks: Development of high-fidelity multi-agent simulators and open-access datasets specifically for CNI applications.
MAS Theory: Coordination, communication protocols, and control of distributed agents within complex, interconnected infrastructure networks.
Grand Challenges: Position papers on open problems where AI/MAS is poised to provide feasible solutions for long-term CNI sustainability.
5. Interactive and Participatory Formats
Technology Demos: Software products, experimental implementations, or multi-agent frameworks designed for CNI applications.
Mini-Tutorials: Concise sessions on AI techniques, AI Safety protocols, or agent-based modeling relevant to critical systems.
Panel Discussions: Moderated debates on emerging issues (e.g., the ethics of autonomous defense or the challenges of certifying CNI agents).
Important dates (EoD Anywhere on Earth (UTC-12))
Call for papers open December 17, 2025
Submission Deadline February 19, 2026
Author Notification March 20, 2026
Camera-ready paper April 24, 2026
Workshop May 26, 2026
Paper categories
Full Paper: Papers with a maximum length of 8 pages plus references (original research)
Work in Progress: Extended abstract with a maximum length of 4 pages plus references (preliminary results, position papers).
Highlight Paper: Summary of previously published papers with a maximum length of 1 page including references that have been published or accepted for publication in archival venues but are relevant for this dedicated community.
Demo: Demonstrations of relevant products, implementations, frameworks, etc. Submit 1 page in the workshop format describing what will be demonstrated and how it fits to the workshop topic.
Mini-Tutorial: Short tutorials in the subject of the workshop, Submit 1 page in the workshop format describing what the tutorial is about, who will be giving it, how it fits to the workshop topic, and the infrastructure necessary (e.g., will participants be expected to have a laptop with installed software to follow the tutorial?).
Formatting guidelines
Please use the following template for your submission:
Submission
Submissions are double-blind (identity or institutions of authors should not be shown or implied in the paper). Make sure you are following the formatting guidelines above and are following the page limitation for your paper category.
Organization
Felipe Leno da Silva - Lawrence Livermore National Laboratory
Wencong Su - University of Michigan - Dearborn
Ruben Glatt - Lawrence Livermore National Laboratory
Christian Schroeder de Witt - University of Oxford
Contact
Email to workshop chairs: ai4cni@gmail.com
Previous workshops
2024: AI4CI @ IJCAI 2024, Jeju Island, Korea
Program Committee
TBD
Page updated
Google Sites
Report abuse