2,529+ academic papers on AI security, attacks, and defenses
Showing 161–180 of 1,906 papers
Clear filtersXinhe Wang, Katia Sycara, Yaqi Xie
Large (vision-)language models exhibit remarkable capability but remain highly susceptible to jailbreaking. Existing safety training approaches aim...
Eungyu Woo, Yooshin Kim, Wonje Heo +1 more
Industrial Control Systems (ICS) integrate computing, physical processes, and communication to operate critical infrastructures such as power grids,...
Sijia Li, Min Gao, Zongwei Wang +3 more
Sequential recommendation seeks to model the evolution of user interests by capturing temporal user intent and item-level transition patterns....
Jiaqi Li, Yang Zhao, Bin Sun +3 more
Autonomous AI agents deployed on platforms such as OpenClaw face prompt injection, memory poisoning, supply-chain attacks, and social engineering,...
Priyal Deep, Shane Emmons, Amy Fox +3 more
LLM-powered applications routinely embed secrets in system prompts, yet models can be tricked into revealing them. We built an adaptive attacker that...
Kato Mivule
This paper extends the Classification Error Gauge (x-CEG) framework, originally developed for measuring the privacy-utility trade-off in tabular...
Qi Li, Bo Yin, Weiqi Huang +6 more
Vision-Language-Action (VLA) models are emerging as a unified substrate for embodied intelligence. This shift raises a new class of safety...
Yu Cui, Ruiqing Yue, Hang Fu +6 more
With the wide adoption of personal AI assistants such as OpenClaw, privacy leakage in user interaction contexts with large language model (LLM)...
Rong Xiang
Recent evidence suggests that frontier AI systems can exhibit agentic misalignment, generating and executing harmful actions derived from internally...
Yuchuan Zhao, Tong Chen, Junliang Yu +3 more
Large language model-powered sequential recommender systems (LLM-SRSs) have recently demonstrated remarkable performance, enabling recommendations...
Jialiang Wang, Yuchen Liu, Hang Xu +7 more
The volume of scientific submissions continues to climb, outpacing the capacity of qualified human referees and stretching editorial timelines. At...
Pegah Khayatan, Jayneel Parekh, Arnaud Dapogny +3 more
Despite impressive progress in capabilities of large vision-language models (LVLMs), these systems remain vulnerable to hallucinations, i.e., outputs...
Naheed Rayhan, Sohely Jahan
Large language models (LLMs) are increasingly integrated into sensitive workflows, raising the stakes for adversarial robustness and safety. This...
Zihan Wang, Rui Zhang, Yu Liu +4 more
LLM agents increasingly rely on skills to encapsulate reusable capabilities via progressively disclosed instructions. High-quality skills inject...
Jiali Wei, Ming Fan, Guoheng Sun +3 more
The growing application of large language models (LLMs) in safety-critical domains has raised urgent concerns about their security. Many recent...
Run Hao, Zhuoran Tan
Model Context Protocol (MCP) is increasingly adopted for tool-integrated LLM agents, but its multi-layer design and third-party server ecosystem...
Arthur Douillard, Keith Rush, Yani Donchev +14 more
Modern large-scale language model pre-training relies heavily on the single program multiple data (SPMD) paradigm, which requires tight coupling...
Yuchen Shi, Xin Guo, Huajie Chen +3 more
Poisoning-based backdoor attacks pose significant threats to deep neural networks by embedding triggers in training data, causing models to...
Vishal Rajput
We prove that empirical risk minimisation (ERM) imposes a necessary geometric constraint on learned representations: any encoder that minimises...
Yongcan Yu, Lingxiao He, Jian Liang +5 more
Test-time reinforcement learning (TTRL) always adapts models at inference time via pseudo-labeling, leaving it vulnerable to spurious optimization...
Get breaking CVE alerts, compliance reports (ISO 42001, EU AI Act), and CISO risk assessments for your AI/ML stack.
Start 14-Day Free Trial
AI Threat Alert