SafeHarness: Lifecycle-Integrated Security Architecture for LLM-based Agent Deployment
Institute of Information Engineering, Chinese Academy of Sciences(中国科学院信息工程研究所);Academy of Mathematics and Systems Science, Chinese Academy of Sciences(中国科学院数学与系统科学研究院);Institute of Applied Physics and Computational Mathematics(应用物理与计算数学研究所);Peking University(北京大学)
Xixun Lin, Yang Liu, Yancheng Chen, Yongxuan Wu, Yucheng Ning, Yilong Liu, Nan Sun, Shun Zhang, Bin Chong, Chuan Zhou, Yanan Cao
The performance of large language model (LLM) agents depends critically on the execution harness, the system layer that orchestrates tool use, context management, and state persistence. Yet this same architectural centrality makes the harness a high-value attack surface: a single compromise at the harness level can cascade through the entire execution pipeline. We observe that existing security approaches suffer from structural mismatch, leaving them blind to harness-internal state and unable to coordinate across the different phases of agent operation. In this paper, we introduce \safeharness{}, a security architecture in which four proposed defense layers are woven directly into the agent lifecycle to address above significant limitations: adversarial context filtering at input processing, tiered causal verification at decision making, privilege-separated tool control at action execution, and safe rollback with adaptive degradation at state update. The proposed cross-layer mechanisms tie these layers together, escalating verification rigor, triggering rollbacks, and tightening tool privileges whenever sustained anomalies are detected. We evaluate \safeharness{} on benchmark datasets across diverse harness configurations, comparing against four security baselines under five attack scenarios spanning six threat categories. Compared to the unprotected baseline, \safeharness{} achieves an average reduction of approximately 38\% in UBR and 42\% in ASR, substantially lowering both the unsafe behavior rate and the attack success rate while preserving core task utility.
CASCADE: Case-Based Continual Adaptation for Large Language Models During Deployment
School of Artificial Intelligence, Jilin University(吉林大学人工智能学院);Engineering Research Center of Knowledge-Driven Human-Machine Intelligence, Jilin University(吉林大学知识驱动人机智能工程研究中心);International Center of Future Science, Jilin University(吉林大学未来科学国际中心);Department of Informatics, King’s College London(伦敦国王学院信息学院);The Alan Turing Institute(艾伦·图灵研究所);AI Centre, Department of Computer Science, UCL(UCL计算机科学系人工智能中心)
Beyond Benchmarks: LLM Evaluation with an Anthropomorphic and Lifecycle-oriented Roadmap
Department of Networks, China Mobile Communications Group Co.,Ltd.(中国移动通信集团有限公司网络部);Xidian University(西安电子科技大学);Oklahoma State University(俄克拉荷马州立大学);Shanghai Jiao Tong University(上海交通大学)
SHE: Trajectory-driven Safety Harness Evolution for LLM Agents
Shanghai Artificial Intelligence Laboratory(上海人工智能实验室);Fudan University(复旦大学);Shanghai Jiao Tong University(上海交通大学);The Hong Kong University of Science and Technology(香港科技大学)
A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deployment
Nanyang Technological University(南洋理工大学);National University of Singapore(新加坡国立大学);The Hong Kong Polytechnic University(香港理工大学);A*STAR(科技研究局);Southern University of Science and Technology(南方科技大学);University of Science and Technology of China(中国科学技术大学);The Pennsylvania State University(宾夕法尼亚州立大学);TeleAI;Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州));Zhejiang University(浙江大学);Institute of Information Engineering, Chinese Academy of Sciences(中国科学院信息工程研究所);Renmin University of China(中国人民大学);University of California, San Diego(加州大学圣地亚哥分校);Tencent(腾讯);Georgia Institute of Technology(佐治亚理工学院);Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所)
Building the ethical AI framework of the future: from philosophy to practice
Graduate School of Global Studies(全球研究研究生院);Tokyo University of Foreign Studies(东京外国语大学);AI Product Development Division(人工智能产品开发部门);Money Forward, Inc.(Money Forward公司)
LLM-Based Data Science Agents: A Survey of Capabilities, Challenges, and Future Directions
York University(约克大学);Vector Institute for AI(人工智能矢量研究所);Dialpad Inc.(Dialpad公司);Nanyang Technological University(南洋理工大学);Salesforce AI Research(Salesforce人工智能研究)
EBCC: Enclave-Backed Confidential Containers via OCI-Compatible Runtime Integration
School of Computer Science and Technology, Xidian University(西安电子科技大学计算机科学与技术学院);Shaanxi Key Laboratory of Network and System Security(陕西省网络与系统安全重点实验室);College of Cyber Security, Jinan University(暨南大学网络空间安全学院)
EBCC 提出了一种 OCI 兼容的运行时架构,用于管理复合保密计算工作负载,通过统一容器化保密计算复合体,保持 OCI 生命周期操作,并将 TEE 特定执行置于后端适配器后。
Adversarial Attacks for Good: A Survey of Proactive Protection across the Visual Content Lifecycle
College of Computing and Data Science, Nanyang Technological University(南洋理工大学计算与数据科学学院);School of Computing and Information Systems, University of Melbourne(墨尔本大学计算与信息系统学院);School of Computing and Information Systems, Singapore Management University(新加坡管理大学计算与信息系统学院);Institute of Trustworthy Embodied AI, Fudan University(复旦大学可信具身人工智能研究院);Department of Computer Science, School of Computing, Institute of Science Tokyo(东京科学大学计算学院计算机科学系);Sony AI(索尼人工智能公司)