Workshop on Trustworthy and Secure AI for Cyber Defense
In conjunction with IEEE DSC 2026

-Introduction
*************************************************
Artificial Intelligence (AI) is transforming cybersecurity by enabling advanced threat detection, vulnerability analysis, incident response, and cyber defense automation. Simultaneously, AI systems themselves have become critical attack surfaces exposed to prompt injection, jailbreak attacks, adversarial manipulation, model poisoning, and other emerging threats. As AI technologies are increasingly deployed in security-critical environments, ensuring their trustworthiness, transparency, robustness, and explainability has become equally important.
The Workshop on Trustworthy and Secure AI for Cyber Defense 2026 aims to bring together researchers, practitioners, industry experts, and policymakers working at the intersection of cybersecurity and artificial intelligence. The workshop provides a forum for discussing advances in AI-driven cyber defense, security evaluation of AI systems, trustworthy and explainable AI, benchmarking methodologies, standardization efforts, and emerging challenges in securing future AI-enabled cyber ecosystems.
This workshop welcomes original research papers, industrial experiences, benchmark studies, datasets, system demonstrations, surveys, and standardization reports that contribute to the development of secure, trustworthy, and effective AI technologies for cybersecurity.
*************************************************

-Topics of Interest
*************************************************
The workshop focuses on advancing research and practice in the following three complementary areas:

1. AI for Cyber Defense
Research on the application of artificial intelligence, machine learning, and foundation models to enhance cybersecurity capabilities, including threat detection, vulnerability analysis, incident response, and cyber resilience.
Topics include, but are not limited to:
•AI-assisted vulnerability detection and analysis
•Benchmarking of LLMs for vulnerability discovery and assessment
•LLM-based software security analysis
•AI-driven malware analysis and classification
•AI-powered threat intelligence
•Autonomous cyber defense agents
•AI-enhanced Security Operations Centers (SOC)
•AI-assisted fuzzing and vulnerability discovery
•AI for incident response and cyber resilience

2. Security of AI Systems
Research on identifying, evaluating, and mitigating security risks in AI systems, foundation models, and AI-enabled applications.
Topics include, but are not limited to:
•Prompt injection attacks and defenses
•Jailbreak attacks against LLMs
•Adversarial machine learning
•Model poisoning and backdoor attacks
•Malicious prompt generation and attack automation
•AI agent security
•Secure deployment of foundation models
•LLM security evaluation methodologies
•Security benchmarking of LLMs and AI agents
•Red teaming for generative AI systems
•Risk assessment frameworks for AI systems

3. Trustworthy AI
Research on improving the transparency, interpretability, robustness, reliability, and human-centered design of AI systems for cybersecurity and other security-critical applications.

Topics include, but are not limited to:
•Explainable AI (XAI) for cybersecurity
•Human-centered explainability
•Uncertainty-aware AI systems
•Human-AI collaboration in security operations
•Trustworthy AI frameworks
•Interpretable machine learning for cyber defense
•Fairness, accountability, and transparency
•Privacy-preserving AI
•AI assurance and certification

In addition, the workshop welcomes contributions on evaluation methodologies, benchmarking, surveys, datasets, governance, and standardization activities related to the above areas, including initiatives from ITU-T, ISO/IEC, ETSI, NIST, and other relevant organizations.
*************************************************

-Paper submission
*************************************************
The workshop welcomes submissions in the following categories.

Full Papers (up to 6 pages): Full papers may present original research contributions, preliminary results, work-in-progress studies, industrial experiences, benchmark studies, surveys, standardization activities, system demonstrations, datasets, and emerging ideas.

Poster Papers (up to 2 pages): Poster papers provide an opportunity to present novel concepts, early-stage research, ongoing projects, and discussion-oriented work.

Papers must be written in English and should not exceed 6 pages for technical papers and 2 pages for poster papers, including figures, tables, references, and appendices. Manuscripts must follow the IEEE Conference Proceedings format (two-column layout, single-spaced, 10-point font). Submitted work must be original and must not have been previously published, nor be under review or consideration for publication elsewhere. All submissions will undergo a peer-review process.

Paper templates can be downloaded from the IEEE website: https://www.ieee.org/conferences/publishing/templates.html
Submission site: https://edas.info/newPaper.php?c=35273&track=139092
Further information or questions may be addressed to: ieeedsc2026-workshop@ml.nict.go.jp

Accepted papers will be presented as part of the IEEE DSC 2026 workshop program. Publication and indexing details will follow the official IEEE DSC 2026 Conference Proceedings policy. The conference proceedings will be published in the IEEE XPlore Digital Library. At least one author of each accepted paper must register for the conference in order for the paper to be included in the proceedings.
*************************************************

-Important Dates
*************************************************
Workshop Paper Submission Deadline: July 17, 2026
Notification of Acceptance: August 7, 2026
Camera-Ready Submission: September 4, 2026
Conference Dates: October 9–11, 2026
*************************************************

-Organizers
*************************************************
Workshop Co-chairs
•Takeshi Takahashi (National Institute of Information and Communications Technology, Japan)
•Gregory Blanc (Telecom SudParis, Institut Polytechnique de Paris, France)

Publicity Chair
•Keisuke Furumoto (National Institute of Information and Communications Technology, Japan)
*************************************************