Topical Area 3: Security of AI and AI for Security
(SAIS)

Riga, Latvia, 23-26 August, 2026

Artificial intelligence is increasingly embedded in decision-making processes, autonomous systems, and data-driven services, creating new security challenges that go beyond those of traditional information systems. Ensuring the integrity, robustness, and dependability of AI-based methods, while exploiting their capabilities to strengthen security mechanisms, has become a central research concern.

At the same time, the accelerated deployment and growing complexity of network technologies, Internet of Things (IoT), and worldwide transformation towards Smart Cities introduce significant challenges in network design, management, and operation. These challenges span hardware and software architecture, service provisioning, interoperability, scalability, and performance optimization. As agentic solutions and applications continue to evolve at a rapid pace, the need for advanced methodologies to manage complexity, ensure reliability, and support intelligent functionality becomes increasingly critical.

The topical area Security of AI and AI for Security focus on advances in protecting AI technologies themselves and on the principled use of AI techniques to enhance security mechanisms. It provides a forum for presenting theoretical foundations, methodological innovations, and experimental results that address security challenges intrinsic to AI models, learning processes, and AI-driven decision systems. This Topical Area is also an interdisciplinary platform for researchers, academics, and practitioners to present and discuss recent advances in the theory, design, and deployment of intelligent, interconnected networked systems, IoT, and Smart Cities, with a particular focus on their security and related aspects of AI.

Topics

Security of Artificial Intelligence

Threat models and vulnerabilities specific to AI and machine learning systems
Adversarial learning techniques and defenses
Robustness, resilience, and reliability of learning algorithms and AI pipelines
Security of data, models, and inference processes across the AI lifecycle
Explainability, transparency, and trustworthiness of AI-based decisions
Privacy risks inherent to AI models and learning processes
Secure design and deployment of AI in safety- and mission-critical contexts

Artificial Intelligence for Security

AI-based methods for detection of misuse, anomalies, and malicious behavior
Learning-based analysis of malware and complex attack patterns
Adaptive and autonomous security mechanisms driven by AI
AI-supported threat analysis and security decision-making
Intelligent incident response and security automation

Cross-Cutting and Emerging Topics

Ethical, legal, and governance challenges related to AI security
Human–AI interaction and decision support in security-critical settings
Evaluation methodologies, benchmarking, and validation of AI-based security solutions
Experimental studies, prototypes, and applied case reports

This topical area complements dedicated thematic sessions focusing specifically on Internet of Things technologies and cybersecurity, privacy, and trust, while providing a unifying framework for broader research on intelligent, scalable, and resilient networked systems. The Security of AI and AI for Security topical area welcomes original research contributions that deepen understanding of AI-specific security risks and demonstrate how AI can be effectively leveraged to strengthen security mechanisms. Submissions that advance foundational theory, propose novel methods, or provide rigorous experimental validation are particularly encouraged.

Topical Area Curators

Armando, Alessandro, University of Genova, Italy
Furtak, Janusz, Military University of Technology, Poland
Suri, Niranjan, Institute of Human and Machine Cognition, United States

Contact: sais@fedcsis.org

Related Thematic Sessions

The following Thematic Tracks are related to this Topical Area:

Agentic AI in SC –
Agentic AI in Smart Cities
IoT‑ECAW –
Internet of Things – Enablers, Challenges and Applications
NEMESIS –
International Forum on Cyber Security, Privacy, and Trust

Submission rules

Authors should submit their papers as Postscript, PDF or MSWord files.
The total length of a paper should not exceed 12 pages IEEE style (including tables, figures and references). More pages can be added, for an additional fee. IEEE style templates are available here.
Papers will be refereed and accepted on the basis of their scientific merit and relevance to the Topical Area.
Preprints containing accepted papers will be published online.
Only papers presented at the conference will be published in Conference Proceedings and submitted for inclusion in the IEEE Xplore® database.
Conference proceedings will be published in a volume with ISBN, ISSN and DOI numbers and posted at the conference WWW site.
Conference proceedings will be submitted for indexation according to information here.
Organizers reserve right to move accepted papers between FedCSIS Sessions.