IET Information Security 

https://ietresearch.onlinelibrary.wiley.com/doi/toc/10.1049/IETIS.si.380506

Artificial Intelligence (AI) techniques, e.g., deep neural networks (DNNs), have been significantly refreshing the state-of-the-art performance of various tasks, e.g., face recognition, robotic grasping, obstacle avoidance of autonomous-driving cars, and UAV Guidance and Planning.

However, recent works report that DNNs are vulnerable to new attack threats such as adversarial attacks and poisoning attacks. Large language models (LLM), which have been popular lately, also face new types of security threats. Therefore, applying these applications to real-world scenarios introduces new security threats and hazards.

This Special Issue aims to explore the security challenges to emerging AI driven applications, such as designing different adversarial attacks and poisoning attacks methods, finding solutions to defend these adversarial attacks and poisoning attacks, extending and investigating adversarial attacks and poisoning attacks to different modalities, different areas and different real-world scenarios. We hope this Issue could welcome original research and review article about the security challenges to emerging AI driven applications and promote building robust AI models.

Potential topics include but are not limited to the following:

• Emerging security threats for 3D point clouds-based application
• Security defense methods for 3D point clouds-based application
• Emerging security threats for 3D mesh-based application
• Security defense methods for 3D mesh-based application
• Emerging security threats for AI driven UAV
• Security defense methods for AI driven UAV
• Emerging security threats for AI driven social networks
• Security defense methods for AI driven social networks
• Emerging security threats for AI driven critical infrastructures
• Security defense methods for AI driven critical infrastructures
• Emerging security threats for LLMs
• Security defense methods for LLMs