arXiv cs.AI· Junchi Liao·· 12 小时前AI 评分31
LLM 安全监控失效：框架边界演进中的证据路径断裂
Monitor-Validity Failures: LLM Safety Evidence Paths at Evolving Framework Boundaries
AI 导读
研究用 canary contracts 协议在公开框架修复中确认了 18 例证据路径失效，覆盖 11 个 LLM 或智能体安全项目组、4 个实现生态。其中 9 例属既有边界修复、8 例为新表面策略覆盖缺口、1 例为纯边界证据问题，14 例触及 E1 执行边界。这些案例表明，即使正则检测器已返回命中，框架仍可能在阻断决策前丢弃该证据，普通应用行为不受影响而安全证据路径已错过决策点。
正文 · 原文
Abstract:Modern LLM applications rely on runtime monitors that inspect tool calls, traces, streams, detector outputs, and verifier inputs. This creates an interface-reliability problem: a framework path may remain operational while policy-relevant evidence never reaches its decision point. In one representative case, a regex detector returned a positive match, but the framework stranded the result before the block decision. This differs from jailbreaks and detector errors because the evidence is already recognized yet disconnected from enforcement. We study this repair-boundary signature with canary contracts: parent/fixed confirmation units that place a known payload only on the changed evidence surface and record monitor visibility together with the downstream decision for action-boundary cases. Applying the protocol to public framework repairs, we confirm 18 evidence-path failures across 11 LLM or agent safety project groups and four implementation ecosystems. Nine are established-boundary repairs, eight are new-surface policy-coverage gaps, and one is a boundary-only evidence case. Fourteen reach E1 enforcement boundaries: seven native action outcomes, six configured enforcement boundaries, and one synthetic-consumer consequence. The remaining cases affect verifier, safety-prompt, or observability channels. The cases form carrier, adapter, authority, and surface families and show how ordinary application behavior can remain intact while a safety-specific evidence path misses its decision. Exact revisions, deterministic outcomes, negative controls, and a frozen prospective audit document the confirmation rule. Monitor assumptions therefore need regression tests at the schemas, streams, tools, and adapters that frameworks evolve.
| Subjects: | Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Software Engineering (cs.SE) |
| Cite as: | arXiv:2610.23362 [cs.AI] |
| (or arXiv:2610.23362v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.23362 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Junchi Liao [view email]
[v1]
Fri, 16 Oct 2026 12:09:26 UTC (276 KB)
来源：arXiv cs.AI · arxiv.org