Entity · paper

Token-Flow Firewall: Semantic Runtime Auditing for Persistent AI Agents

paperactiveprovisionaltoken-flow-firewall-semantic-runtime-auditing-for-persistent-ai-agents-b31e93db·1 events·first seen 3d ago

Aliases: Token-Flow Firewall: Semantic Runtime Auditing for Persistent AI Agents

Co-occurring entities

TokenWall CIK-Bench

More like this (12)

The Unfireable Safety Kernel: Execution-Time AI Alignment for AI Agents and Other Escapable AI Systems LedgerAgent: Structured State for Policy-Adherent Tool-Calling Agents Efficient and Sound Probabilistic Verification for AI Agents Can Trustless Agents Be Trusted? An Empirical Study of the ERC-8004 Decentralized AI Agent Ecosystem Always-OnAgents: A Survey of Persistent Memory, State, and Governance in LLM Agents Trustworthy AI Linguistic Firewall: Geometry as Defense in Multi-Agent Systems Routing Beyond Runtime Enforcement: Shield Synthesis as Defensibility Analysis for Adversarial Networks Criticality-Based Guard Rail Validation for AI Agent Decisions in Autonomous Telecom Networks Privacy-Preserving RAG via Multi-Agent Semantic Rewriting Learning Red Agent Policy from Observations for Neurosymbolic Autonomous Cyber Agents Concrete Problems in AI Safety

Recent events (1)

6arXiv · cs.CL·3d ago·source ↗

TokenWall: Semantic runtime firewall for persistent AI agent token flows

Researchers introduce TokenWall, a runtime defense framework that audits natural-language token flows in persistent AI agents before they reach privileged execution sinks. The system constructs source-sink audit records, applies lightweight local inspection, and escalates ambiguous cases to stronger arbitration modules. On CIK-Bench, TokenWall reduces attack success rate to 12.5% while maintaining a 97.4% benign pass rate and adding only 0.69 seconds of latency, demonstrating a practical security-utility tradeoff for long-lived agentic systems.

AI Safety Research Agent and Tool Ecosystem Token-Flow Firewall: Semantic Runtime Auditing for Persistent AI Agents TokenWall CIK-Bench