Skip to content
Preprint

Containing the Autonomous Operator: A Defense-in-Depth Framework and Reference Architecture for Securing AI Agents on Kubernetes

Oct 2026 · 0 citations · 42 references
Computer Science

TL;DR

This paper argues that the model must not be treated as a security boundary and that agent safety on Kubernetes is therefore an infrastructure problem: every guarantee must continue to hold under the assumption that the agent is fully compromised by prompt injection.

Abstract

Large language model (LLM) agents are moving from chat interfaces into infrastructure operations, where they read telemetry, call tools, generate and execute code, and change the state of production Kubernetes clusters. This collapses a boundary that conventional cloud-native security assumes: the boundary between data and control. Content that an agent merely reads (a log line, a ticket, a tool description) can redirect what it does. This paper argues that the model must not be treated as a security boundary and that agent safety on Kubernetes is therefore an infrastructure problem: every guarantee must continue to hold under the assumption that the agent is fully compromised by prompt injection. We contribute (i) a threat model and ten-class threat taxonomy for agents operating on and within Kubernetes, aligned with emerging OWASP guidance for agentic applications; (ii) nine design principles, centered on complete mediation at the tool boundary and on breaking the combination of untrusted input, sensitive access, and external egress; (iii) a seven-layer defense-in-depth framework that maps each principle to native or widely adopted Kubernetes mechanisms: workload identity, RBAC and ValidatingAdmissionPolicy, gVisor/Kata sandboxing via the SIG Apps Agent Sandbox project, FQDN-aware egress policy, an agent/MCP gateway with policy-as-code over tool arguments, and eBPF runtime enforcement; (iv) a reference architecture with concrete policy artifacts and per-layer bindings for Amazon EKS, Azure Kubernetes Service, and Google Kubernetes Engine; and (v) a qualitative evaluation comprising a threat-control coverage matrix and four attack walkthroughs, with a proposed empirical methodology. We report no measured attack-success or overhead figures; instead we identify residual risks and the measurements needed to validate the framework.

View source

Similar papers

#machine learning Preprint Oct 2026

Kairos: Transactional Security Enforcement at the System Boundary for LLM Agents

Kairos is presented, a reference monitor at the isolation boundary of an unmodified, opaque agent, where a tool call, a shell command, and a script the agent writes all cross the boundary as the same file and network operations, so that confinement holds regardless of the agent's behavior.

Kristoffer Hempel, Nils Loose, Lorenzo Cavallaro et al. · 0 citations

When the Agent Becomes the Kernel: A Systematization of Security on the Path to AI-Native Operating Systems

It is argued that this gap leaves an irreducible residual of undetected attacks wherever inputs and actions are not restricted in advance to an enumerated set, and systematize defenses across runtime monitoring, architectural separation, and authorization.

Li Zhang, Yang Sun, Jie Shi · 1 citation
#explainable ai Review Open access Oct 2026

Agentic AI-Driven SOC-as-a-Service for Optimizing Incident Response in Cloud Environments: A Conceptual Framework

Cloud security operations must process high-volume, rapidly changing telemetry within short response windows while maintaining governance and accountability. However, the literature on cloud defense, SOC modernization, retrieval-augmented generation (RAG), and Agentic AI remains fragmented. This paper combines a legacy...

Abdulaziz Y. Alhumaidi, Faisal A. Al-Qadda, Albandari Alsumayt et al. · 0 citations

Hard Stop: Kernel-Level Preemption and Containment for Rogue Agentic Execution

This monograph presents a first-principles forensic autopsy of the intrusion, provides formal evidence that the breach was a predicted consequence under the Instrumental Convergence thesis operating within an unattenuated autonomous loop lacking out-of-band circuit-breakers, exposes the Defensive LLM Guardrail Paradox...

José Luis Pino · 0 citations
#artificial intelligence Preprint Oct 2026

Multi-Aspect Runtime Verification for Simulation-Based V&V of LLM-Enabled Autonomous Agents

A multi-aspect runtime-verification framework that decomposes a natural-language policy clause into a typed spatial/temporal/semantic triple over one canonical event stream, checks each aspect with its own monitoring specification, and fuses the verdicts through a four-valued algebra that carries provenance.

Nikolaos Kekatos, Dimitrios Nikou, Anastasios Temperekidis et al. · 0 citations

Ajar: Measuring Open Privilege in Agent Defenses

Ajar is evaluated by attaching it to AgentDojo, where open privilege becomes a third axis beside the existing attack success and benign utility, and it is observed that they leave widely different amounts of privilege open.

Reshabh K Sharma, Lin-Xi Jiang, Shuo Chen et al. · 1 citation

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.