CORTEXA
← Browse
arxivcs.CRcs.AI2026-07-11

Temporary Authority, Permanent Effects: Commit-Time Authorization for LLM Agents

Igor Santos-Grueiro

LLM agents can commit durable effects from authority evidence that was valid earlier in execution: a DOM snapshot, approval epoch, version witness, branch token, or worker result. We study the commit boundary at which earlier authority evidence no longer authorizes a durable effect. We call this property commit-time authorization: a durable effect is authorized only if the witness that licensed its derived state remains fresh, causally prior, bound to the same effect, and eligible at commit time. We build a controlled-invalidation suite spanning browser, tool/API, and multi-agent workflows. The suite preserves the user goal and payload shape while invalidating the authority relation before durability. In the primary 54-task matrix, endpoint success remains high: 262/270 runs reach the visible result. Only 55/270 are authorized completions; among the 216 invalidating rows, 207 commit after the authorizing path has failed. All 54 clean controls remain authorized, and a separate 54-run authority-preserving check produces no unauthorized commits. We then evaluate mitigation families. Prompt caution and single-condition checks are insufficient because different hazards break different boundary conditions. Defenses work when they refresh, rebind, replan, or refuse at the durability boundary. CommitGuard, a fail-closed boundary monitor, blocks stale durable-effect attempts on protected commit surfaces when runtimes emit witness, dependency, binding, and eligibility signals. The result is a reporting and runtime-design lesson: endpoint success is a utility metric; authorized commit is a security property.

View free PDFSource page

Related papers

arxivcs.CRcs.AI2026-06-26

LLM agents security duality: a comprehensive survey of self-security and empowered cybersecurity

Yiwei Xu, Yong Zhuang, Xuanming Liu, Tian Zhang, Bowen Xiao, Xiaoyang Xu, et al.

Large language model (LLM) agents are rapidly being integrated into real-world systems. Their autonomy and tool-use capabilities generate substantial value while simultaneously expanding the security attack surface. This survey provides a comprehensive overview of the opportuniti…

View free PDFSource page
arxivcs.CRcs.AIcs.LG2026-07-20

Adaptive Adversaries: A Multi-Turn, Multi-LLM Benchmark for LLM Agent Security

Devina Jain, David Hartmann, Chuan Li

LLM-based agents process external content, exposing them to prompt injection and multi-turn manipulation. Most safety benchmarks evaluate defenders against fixed attack pools collected before evaluation, single-turn or multi-turn. We present a 21-scenario benchmark for \emph{adap…

View free PDFSource page
arxivcs.CRcs.AI2026-07-21

Cross-Agent Campaign Attribution: Linking Asynchronous Attacks Across LLM Agents

SangJin Park, Myungsub Choi, Jineok Kim, Minseung Kang

LLM-agent defenses are typically evaluated one session at a time. In deployment, however, attacks can be distributed across independent agents, teams, and runtimes, leaving each local guardrail with only a sparse fragment. We formalize cross-agent asynchronous campaign attributio…

View free PDFSource page
arxivcs.CRcs.AIcs.MA2026-07-21

Broken Gates: Re-evaluating Web Bot Defenses in the Age of LLM Agents

Behzad Ousat, Nikita Turkmen, Lalchandra Rampersaud, Dillan Bailey, Amin Kharraz

LLM-based browser agents are rapidly changing the threat landscape for web security. Unlike traditional automation frameworks that execute predefined scripts, these agents can autonomously navigate websites, reason about page content, and interact with web interfaces using natura…

View free PDFSource page
arxivcs.CRcs.AIcs.LG2026-07-09

TRACE: A Two-Channel Robust Attribution Watermark via Complementary Embeddings for LLM-Agent Trajectories

Zheng Gao, Xiaoyu Li, Xiaoyan Feng, Jiaojiao Jiang, Yang Song, Yulei Sui, et al.

LLM agents reach users through resellers, who may rebrand a developer's agent or substitute a cheaper model. When provenance is disputed, attribution rests on the trajectory log (the record of tool calls, observations, and executed actions, not the model's reasoning), which the r…

View free PDFSource page