Important dates
- Submission deadline: 15 January 2027
- First-round reviews to authors: 15 April 2027
- Revised manuscripts due: 15 June 2027
- Second-round reviews / notification: 31 August 2027
- Final (camera-ready) manuscripts due: 15 October 2027
- Expected publication: Early 2028
Intelligent systems built on foundation models and autonomous agents are being embedded into the infrastructures that underpin our digital lives. These systems plan, retrieve, and act with growing autonomy, and the move from intelligent to agentic computing introduces new attack surfaces — prompt injection, tool and function-call abuse, agent collusion, memory and context corruption, and emergent failures — that classical defenses do not address. The same models are also reshaping defense, enabling automated vulnerability discovery, AI-driven detection, and resilience engineering at new scale. This special issue seeks original, high-quality work spanning both directions: securing and governing intelligent and agentic systems, and harnessing AI to advance trust, privacy, and security.
Topics of interest include, but are not limited to:
- Foundation Models and LLMs — trust, safety, privacy, and resilience of large language and (multimodal) foundation models; jailbreaks and prompt injection; data extraction and leakage; RAG security and unlearning.
- Agentic and Multi-Agent AI — tool-use isolation and least-privilege execution; agent identity and delegated access control; multi-agent and protocol security (e.g., MCP); memory/context integrity and human oversight.
- Frontier AI Safety, Assurance, and Governance — alignment and control of highly capable models; auditing, red-teaming, and evaluation; conformance with risk frameworks and regulation (e.g., NIST AI RMF, EU AI Act); fairness and integrity.
- Privacy-Preserving and Cryptographic ML — differential privacy; federated and decentralized learning; confidential computing and TEEs; encrypted computation and MPC; synthetic-data governance; post-quantum crypto-agility.
- Adversarial ML and Threat Modeling — adversarial attacks and defenses; threat modeling for AI/ML systems; supply-chain security (poisoned weights, provenance, model cards, AIBOM); risk measurement and mitigation.
- AI for Security and Trust — AI/ML for detecting malware, phishing, and APTs; LLM assisted vulnerability discovery and autonomous defense; dual-use and offensive-AI risk; insider-threat modeling.
- Trustworthy Systems and Infrastructures — foundational models of trust, privacy, and security; resilient intelligent systems; security for IoT, cyber-physical-human, and edge/cloud platforms; trust management and human factors.
- Content Authenticity and Emerging Applications — provenance and watermarking of AI-generated content; deepfakes and social cybersecurity; and trustworthy autonomous/embodied agents, critical infrastructure, smart cities, blockchain, and services computing.
Submission Guidelines:
Papers must be original and not under consideration elsewhere. Submissions that extend a previously published conference paper must contain at least approximately 30% new material beyond the conference version - for example, substantial new methodology, analysis, or experiments; expanded motivation, related work, or incremental experiments alone are insufficient. Authors must cite the prior conference version and attach it at the time of submission and must clearly indicate how the extended manuscript differs from and substantially goes beyond that version. TDSC does not accept survey or case-study papers.
Selection and Evaluation Criteria
- Relevance to the topics of this special issue
- Significance to the scope of IEEE TDSC
- Novelty, technical depth, and potential impact
- Soundness and quality of the contribution
- Clarity and readability
- Adequacy of references and treatment of related work
- Original, first-submission work not published elsewhere
For author information and guidelines on submission criteria, please visit the Author Information Page. Please submit papers through the IEEE Author Portal, and be sure to select the "SI: Trust, Privacy, and Security in Intelligent and Agentic Systems” subject. Manuscripts should be written in English and describe original research. Manuscripts should not be published or currently submitted for publication elsewhere. Please submit only full papers intended for review, not abstracts. The review process will comply with the standard review process of the IEEE Transactions on Dependable and Secure Computing journal. All submitted papers will be evaluated on the basis of relevance, significance of contribution, and technical quality.
Questions?
Contact the guest editors at:
