Topics of interest

• Risk identification and modeling in AI-driven systems and multi-component ecosystems.
• Human–AI interaction risks, trust, overreliance, and decision-making under automation.
• Human-centered cybersecurity and social cyber attacks, including phishing, social engineering, and disinformation.
• Machine learning and knowledge discovery for cyber threat detection on large-scale heterogeneous data.
• Natural language processing and large language models for malicious or manipulative content analysis.
• Graph learning and knowledge graphs for social networks, communication patterns, and cyber threat ecosystems.
• Bias, robustness, uncertainty, and adversarial machine learning in human-critical settings.
• Explainable and trustworthy AI for security-critical applications.
• Generative AI for both cyber attacks and defensive analysis.
• Runtime monitoring, continual risk assessment, and adaptive mitigation strategies.
• Human behavioral and psychophysiological data analysis.
• Privacy, data protection, cyber threat intelligence, and knowledge sharing.
• Human-in-the-loop and reasoning-based methods for dynamic risk assessment and explanation.
• Case studies and empirical evaluations in real-world deployments.

Submission guidelines

Papers must be written in English and formatted in LaTeX following the Springer LNCS author kit. The maximum length is 16 pages including references. Over-length papers will be rejected without review.

Up to 10 MB of additional material may be uploaded with the submission. Appendices should be submitted separately, and the main paper must remain self-contained within the page limit.

SHIELD-AI 2026 welcomes regular research papers, survey articles, mature work, recent research contributions, preliminary exploratory work, and industry experience or case-study papers. Submissions must be original and not under review elsewhere.

All submissions should be made in PDF via Microsoft CMT and must adhere to the Springer LNCS style. At least one author of each accepted paper must register and present the paper in person in Naples.

The workshop papers may be included in a joint post-workshop proceeding published by Springer Communications in Computer and Information Science (CCIS), with authors able to opt in or out.