Airplane Autopilot — agentic threat model
Airplane Autopilot presents a high-risk profile due to its role in building and orchestrating internal corporate tools, where compromise could lead to unauthorized code execution or data exposure across internal systems. The lack of documented sandboxing or guardrails in the public listing amplifies these risks.
OWASP AIVSS score rationale
| Autonomy of Action | 0.50 | |
| Goal-Driven Planning | 0.60 | |
| Self-Modification | 0.10 | |
| Dynamic Tool Use | 0.70 | |
| Persistent Memory | 0.30 | |
| Contextual Awareness | 0.50 | |
| Dynamic Identity | 0.40 | |
| Multi-Agent Interactions | 0.20 | |
| Non-Determinism | 0.60 | |
| Opacity & Reflexivity | 0.50 |
Scored with the canonical OWASP AIVSS formula (AIVSS calculator reference); agentic risk factors estimated from the agent’s described capabilities.
MAESTRO 7-layer threat model
Per-layer threats for this agent. Layers tagged “not certain from listing” are general, caveated commentary where the public description didn’t pin that layer.
Not certain from the listing — The specific foundation models used by Airplane Autopilot are not disclosed. Standard LLM threats like prompt injection, adversarial examples, and misaligned outputs remain a general concern for any internal tool builder.
Not certain from the listing — No details are provided regarding RAG, vector databases, or training data pipelines. If integrated with Airplane's internal tools, data exfiltration or knowledge-base poisoning could expose sensitive corporate data.
Not certain from the listing — As an autopilot for building internal tools, it likely orchestrates workflows and calls APIs. This introduces risks of tool misuse, insecure tool integration, and prompt injection leading to unauthorized actions.
Not certain from the listing — The hosting environment, sandboxing of executed code, and secrets management are not described. Given it builds internal tools, lack of strict sandboxing could lead to remote code execution or privilege escalation.
Not certain from the listing — There is no mention of evaluation frameworks, guardrails, or logging mechanisms to detect anomalous agent behavior or drift.
Not certain from the listing — Compliance certifications (e.g., SOC2) or identity/access management controls are not detailed, which is critical since the agent interacts with internal corporate tools.
Not certain from the listing — It is unclear if this autopilot interacts with other agents or external marketplaces, but multi-agent trust abuse could occur if it integrates with other internal automated systems.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).