aiqbench analysis
This paper introduces a compositional framework that quantifies residual risk in agentic AI by decomposing failure paths across seven physical and logical layers. By grounding control effectiveness in state dynamics rather than informal scores, it offers a rigorous method for assessing trust boundaries in both embodied robots and financial agents without treating internal mechanisms as black boxes. The approach provides a transferable pipeline for building resilient systems where governance observability is additive to the core risk functional.
From Agent Failure Paths to Quantified Residual Risk: A Compositional Framework for Resilient Agentic AI
arXiv:2607.18243v1 Announce Type: new Abstract: Agentic AI is crossing trust boundaries faster than current risk models can represent. Existing approaches provide one of two partial views. They either describe failure mechanisms without producing a transferable residual-risk estimate, or they produce a risk estimate while treating the internal failure path as a black box. We couple those two views by proposing CPSAINT, a seven-layer integrity decomposition over Physical state, Sensors, Data, Compute, Actuators, Environment, and Time, paired with FRIESA-K, a residual-risk functional that maps each failure path to a quantified risk instance. FRIESA-K grounds the resistance term K in a controlled absorbing Markov model so that control effectiveness is derived from state dynamics rather than assigned as an informal score. The result is a concise mechanism-to magnitude pipeline for resilient agentic and embodied AI. We report governance observability through a separate additive penalty instead of inserting governance as a new variable in the resistance functional. We formalize structural composability linking valid failure paths to well-defined risk instances and show the framework on two contrasting scenarios a hard real-time warehouse robot and a governance-instrumented financial-services agent. Across both cases, the same layer grammar, variable semantics, and dynamic-resistance construction remain intact. Thus, we obtain a compact kernel that supports cross-domain reasoning, explicit assumptions, and quantitatively grounded formalism of composable trust.
Abstract:Agentic AI is crossing trust boundaries faster than current risk models can represent. Existing approaches provide one of two partial views. They either describe failure mechanisms without producing a transferable residual-risk estimate, or they produce a risk estimate while treating the internal failure path as a black box. We couple those two views by proposing CPSAINT, a seven-layer integrity decomposition over Physical state, Sensors, Data, Compute, Actuators, Environment, and Time, paired with FRIESA-K, a residual-risk functional that maps each failure path to a quantified risk instance. FRIESA-K grounds the resistance term K in a controlled absorbing Markov model so that control effectiveness is derived from state dynamics rather than assigned as an informal score. The result is a concise mechanism-to magnitude pipeline for resilient agentic and embodied AI. We report governance observability through a separate additive penalty instead of inserting governance as a new variable in the resistance functional. We formalize structural composability linking valid failure paths to well-defined risk instances and show the framework on two contrasting scenarios a hard real-time warehouse robot and a governance-instrumented financial-services agent. Across both cases, the same layer grammar, variable semantics, and dynamic-resistance construction remain intact. Thus, we obtain a compact kernel that supports cross-domain reasoning, explicit assumptions, and quantitatively grounded formalism of composable trust.
Artificial Intelligence (cs.AI)
arXiv:2607.18243 [cs.AI]
(or arXiv:2607.18243v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2607.18243 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 304 words
Related coverage
Improving Bash Generation in Small Language Models with Grammar-Constrained Decoding
NVIDIA · May 8, 2026
High-Throughput Structure Prediction with BioNeMo Inference Runtime
NVIDIA · Sep 10, 2026
The AI - Quantum Convergence Hits Critical Mass
livetradingnews.com · Aug 3, 2026
The standard for agents you can trust: Lessons from the federal front lines
Snorkel AI Blog · Jun 5, 2026