Runtime Safety Filtering

Follow this topic Ask about this topic

Momentum

4 papers in the last four weeks, down 20% on the four weeks before. 0.0% of all new papers.

Jul 13Week of Sep 28

Latest papers 33

All topics

Instrumental Monitor Evasion Emerges Under Ordinary Task Pressure
Sep 24, 2026David Schmotz, Derck Prinzhorn, Luca Beurer-Kellner +3Runtime Safety Filtering Evasion
ELLIS Institute Tübingen · Max Planck Institute for Intelligent Systems · Tübingen AI Center +3
ClashBench: Conflicts Leading Agents to Seize and Harm
Sep 17, 2026Yuejin Xie, Yu Li, Dadi Guo +6Runtime Safety Filtering Preemption
Tsinghua University · Shanghai AI Lab · Fudan University +2
Runtime Safety Filtering for Two-Terminal Hazards in Robotic Battery Recycling
Sep 17, 2026Yuxin Cao, Wei Song, Xianglin Yang +4Safety Filters Runtime Safety Filtering
National University of Singapore, Singapore · Griffith University, Australia · UNSW Canberra, Australia +2
Semantic-TVM: Structure-Preserving Trustworthy Virtual Memory for Memory-Augmented and Tool-Using Agents
Sep 14, 2026Yu Li, Qikun Cai, Tao Huang +1Peak Memory Untrusted Content
School of Computer and Big Data, MinJiang University Fuzhou, China
Defense-as-Skill: Evolving Runtime Guard Skill for Skill-Augmented Agents
Sep 1, 2026Xiaofang Yang, Ziqi Miao, Dianbo Sui +2Runtime Safety Filtering Malicious Agents
Shanghai Artificial Intelligence Laboratory · Harbin Institute of Technology, Weihai
The Irreversibility Budget: Fleet-Level Risk Accounting and Admission Control for Agent Operating Systems
Aug 31, 2026Bardia Mohammadi, Laurent BindschaedlerRuntime Safety Filtering Agentic Deployments
Max Planck Institute for Software Systems Saarbrücken, Germany
SingProbe Technical Report
Aug 31, 2026Sing TeamLarge Language Model Safety Runtime Safety Filtering
AI Security Lab, Ant Group
Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents
Aug 13, 2026Jiajun Ruan, Peiyang Li, Yukun Chen +2Runtime Safety Filtering Large Language Model Agents
University of Minnesota · Ant Group · Tsinghua University +1
Dual Stress: Runtime Safety Monitoring for Safety-Constrained MPC Navigation
Aug 11, 2026Jamil Chahine, Wenqi Cai, John Abanes +1Safety Constraints Model Predictive Control
Department of Mechanical Engineering American University of Beirut (AUB) Beirut 1107 2020, Lebanon · Robotics and Intelligent Systems Control Laboratory New York University Abu Dhabi (NYUAD) Abu Dhabi, UAE · Department of Electrical and Computer Engineering New York University New York, NY, USA +1
DreamGuard: Efficient Runtime Guardrail for LLM Agents via Risk-Aware World Model
Aug 6, 2026Wenhao Lin, Chenyu Yu, Xingwei Lin +6Runtime Safety Filtering Large Language Model Agents
Zhejiang University · Nanjing University of Posts and Telecommunications · Sun Yat-sen University
Why Formal Monitors Fail: Attack Distribution Entropy as a Coverage Bound for LTL-Based LLM Agent Safety
Aug 2, 2026Ruiyang ZhangLarge Language Model Safety Runtime Safety Filtering
Ryonix Labs Inc. Flock.io
ActFovea: Runtime Safeguarding for VLA Policies via Spatiotemporal Visual-Action Consistency
Jul 31, 2026Wenda Yu, Tianshi Wang, Fengling Li +3Robotic Manipulation Runtime Safety Filtering
Tongji University · Mohamed bin Zayed University of Artificial Intelligence · Shanghai Artificial Intelligence Laboratory +1
Democratizing Agent Deployment Safety: A Structural Monitoring Approach
Jul 16, 2026Preeti Ravindra, Rahul Tiwari, Vincent WolowskiAgentic Deployments Security Evaluation
1Independent
Heterogeneous Agent Cohorts for Safe Open-Ended Exploration with Runtime Constraint Memory
Jul 13, 2026Tengjiao LiuRuntime Safety Filtering Runtime
Founder & Researcher, psi.run
Runtime Safety Filtering for Learned Small UAS Separation Policies under GNSS Degradation
Jul 10, 2026Alex Zongo, Peng WeiMulti-Uav Safety Filters
Department of Mechanical and Aerospace Engineering · George Washington University · Washington, D.C., USA
Token-Flow Firewall: Semantic Runtime Auditing for Persistent AI Agents
Jul 9, 2026Puji Wang, Yingchen Zhang, Ruqing Zhang +2Agentic Deployments Semantic Firewall
State Key Laboratory of AI Safety · Institute of Computing Technology, Chinese Academy of Sciences · University of Chinese Academy of Sciences, Beijing, China
EgoSafetyBench: A Diagnostic Egocentric Video Benchmark for Evaluating Embodied VLMs as Runtime Safety Guards
Jun 30, 2026Siddhant Panpatil, Arth Singh, Mijin Koo +3Safety Benchmarks Egocentric Video
AIM Intelligence · Seoul National University
LabGuard: Grounding Natural-Language Laboratory Rules into Runtime Guards for Embodied Laboratory Agents
Jun 30, 2026Jingpu Yang, Fengxian Ji, Zhengzhao Lai +8Runtime Safety Filtering Laboratory
1Wuhan University · 2MBZUAI · 3Northeastern University
Mostly Automatic Translation of Language Interpreters from C to Safe Rust
Jun 25, 2026Bo Wang, Brandon Paulsen, Joey Dodds +3Program Analysis Runtime Safety Filtering
National University of Singapore, Singapore · Amazon, USA
RedAct: Redacting Agent Capability Traces for Procedural Skill Protection
Jun 9, 2026Shuwen Xu, Zhitao He, Yi R. FungTraces Data Leakage
1Hong Kong University of Science and Technology · University of Chinese Academy of Sciences
Agent libOS: A Runtime Substrate for Capability-Controlled Self-Evolving LLM Agents
Jun 2, 2026Yingqi ZhangLarge Language Model Agents Runtime Safety Filtering
Department of Computer Science and Technology, Tsinghua University Beijing, China
GNMR: Runtime Stability Control for Low-Precision Large Language Model Training
May 30, 2026Boao Kong, Weichen Jia, Engao Zhang +6Large Language Model Training Full-Precision Performance
Peking University · Huawei Technologies Ltd.
AIRGuard: Guarding Agent Actions with Runtime Authority Control
May 27, 2026Suliu Qin, Haomin Zhuang, Yujun Zhou +2Authorization Runtime Safety Filtering
University of Liverpool · University of Notre Dame · 2Inria, France
Silent Failures in Physical AI: A Literature Review of Runtime Action Authorization for Autonomous Systems
May 23, 2026Barak OrArtificial Intelligence Safety Runtime Safety Filtering
STATE16
PragLocker: Protecting Agent Intellectual Property in Untrusted Deployments via Non-Portable Prompts
May 7, 2026Qinfeng Li, Yuntai Bao, Jianghui Hu +5Untrusted Content Prompt Engineering
Zhejiang University · Chang’an University · Innovation and Management Center, School of Software Technology (Ningbo), Zhejiang University +2
AgentTrust: Runtime Safety Evaluation and Interception for AI Agent Tool Use
May 6, 2026Chenglin YangRuntime Safety Filtering Untrusted Content
Independent Researcher
Worst-Case Discovery and Runtime Protection for RL-Based Network Controllers
May 6, 2026Hongyu Hè, Minhao Jin, Maria ApostolakiReinforcement Learning Control Worst-Case
Princeton
Safe Bilevel Delegation (SBD): A Formal Framework for Runtime Delegation Safety in Multi-Agent Systems
Apr 30, 2026Yuan SunDelegation Model-Based Multi-Agent Systems
Jilin University
AgentWard: A Lifecycle Security Architecture for Autonomous AI Agents
Apr 27, 2026Yixiang Zhang, Xinhao Deng, Jiaqing Wu +3Runtime Safety Filtering Artificial Intelligence Agents
Layerwise Convergence Fingerprints for Runtime Misbehavior Detection in Large Language Models
Apr 27, 2026Nay Myat Min, Long H. Pham, Jun SunFingerprint Jailbreak Success Rates
Singapore Management University
SafeAgent: A Runtime Protection Architecture for Agentic Systems
Apr 19, 2026Hailin Liu, Eugene Ilyushin, Jie Ni +1Runtime Safety Filtering Large Language Model Safety
Lomonosov Moscow State University · Central University
Evaluating Tool-Using Language Agents: Judge Reliability, Propagation Cascades, and Runtime Mitigation in AgentProp-Bench
Apr 17, 2026Bhaskar GurramMemoryagentbench Runtime Safety Filtering
Zasti Inc., Ashburn, VA, USA
LinuxArena: A Control Setting for AI Agents in Live Production Software Environments
Apr 16, 2026Tyler Tracy, Ram Potham, Nick Kuhn +31Sabotage Runtime Safety Filtering
Redwood Research · EquiStamp

Sep 24, 2026, cs.CRJ/K move · Enter open · S save

Instrumental Monitor Evasion Emerges Under Ordinary Task Pressure

David Schmotz, Derck Prinzhorn, Luca Beurer-Kellner, Anselm Paulus+2

ELLIS Institute Tübingen · Max Planck Institute for Intelligent Systems · Tübingen AI Center+3

Runtime Safety Filtering Evasion Artificial Intelligence Safety Scalable Oversight

Open paper AskPDF arXiv

A central concern in AI safety is that agents may treat oversight as an obstacle when it conflicts with completing their goals. We study instrumental evasion, the propensity of LLM agents to circumvent runtime monitoring as a means of completing ordinary tasks. We introduce EvasionBench, a benchmark of 50 diverse task-policy pairs in which completing the task requires an operation prohibited by a runtime monitor. Agents know that their tool calls are monitored and are prompted to continue working when they pause. Across our evaluations, best-of-3 evasion attempt rates reach up to 98% and success rates up to 88%, with substantial variance across models. Claude Fable 5.1 succeeds less often, but frequently makes creative attempts to circumvent the monitor. Evasion generally increases with test-time compute, with higher evasion rates at greater reasoning effort and token use. Traces show that agents encode prohibited commands, decompose operations across tool calls, and retry until relevant context leaves the monitor's history. We also find that GPT-6 Astra's low evasion rate comes with overrefusal, as it frequently abandons otherwise solvable tasks under a denial-of-service prompt injection. Our findings show that ordinary task pressure can lead to adaptive attempts to evade runtime monitors without an explicit adversarial objective. Effective oversight must therefore remain robust against repeated attempts, as the persistence that helps agents solve difficult tasks can also drive them to circumvent their guardrails.

Momentum

Related topics

Latest papers 33