From Agent Failure Paths to Quantified Residual Risk: A Compositional Framework for Resilient Agentic AI
Quick Answer
This paper shows that The CPSAINT framework integrates failure mechanisms with residual-risk estimates for agentic AI, using a seven-layer integrity model and FRIESA-K for quantifying risks.
Quick Take
This approach enhances resilience in AI systems, demonstrated through contrasting applications in warehouse robotics and financial services, while maintaining a consistent layer grammar and dynamic resistance construction.
Key Points
- CPSAINT features a seven-layer integrity decomposition for agentic AI risk assessment.
- FRIESA-K maps failure paths to quantified risk instances using a controlled Markov model.
- The framework supports cross-domain reasoning and formalizes composable trust.
- Demonstrated effectiveness in both warehouse robotics and financial services scenarios.
- Governance observability is achieved through an additive penalty mechanism.
DeepSignal Analysis
What happened
The CPSAINT framework has been proposed to integrate failure mechanisms with residual-risk estimates for agentic AI. It employs a seven-layer integrity model and the FRIESA-K functional to quantify risks, demonstrated through applications in warehouse robotics and financial services.
Key evidence
- CPSAINT combines failure mechanisms with residual-risk estimates, addressing a gap in existing AI risk models.
- The framework uses a seven-layer integrity model that includes Physical state, Sensors, Data, Compute, Actuators, Environment, and Time.
- FRIESA-K maps failure paths to quantified risk instances, grounding control effectiveness in state dynamics rather than informal scores.
Why it matters
This framework aims to enhance the resilience of AI systems by providing a structured approach to understanding and quantifying risks. By linking failure paths to risk estimates, it offers a more comprehensive view of AI reliability, which is crucial as agentic AI technologies advance rapidly.
What to watch
It will be important to observe how CPSAINT performs in real-world applications beyond the tested scenarios. Additionally, the effectiveness of the governance observability mechanism and its impact on risk management in various domains should be monitored.
Paper Resources
📖 Reader Mode
~2 min readAbstract: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.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2607.18243 [cs.AI] |
| (or arXiv:2607.18243v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2607.18243 arXiv-issued DOI via DataCite |
Submission history
From: Deepti Gupta [view email]
[v1]
Tue, 28 Apr 2026 23:40:24 UTC (294 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.AI
See more →HOBA: Hierarchical On-Policy Bidding Agents for Adaptive Online Advertising
HOBA (Hierarchical On-policy Bidding Agents) is a novel hierarchical reinforcement learning framework that enhances online advertising bidding systems by improving adaptability and reducing hyperparameter tuning costs. It utilizes a for hyperparameter inference, a SARSA agent for expert model selection, and a dynamic expert pool for bid execution, achieving a +3.6% increase in target cost during large-scale deployment and outperforming state-of-the-art baselines on AuctionNet.