Gander — agentic threat model
Gander possesses a high-risk profile due to its direct integration with financial reimbursement systems, booking databases, and telephony networks, enabling automated real-world actions like payouts and rebookings. A compromise could lead to significant financial fraud, unauthorized flight/hotel bookings, and mass exposure of passenger PII.
OWASP AIVSS score rationale
| Autonomy of Action | 0.80 | |
| Goal-Driven Planning | 0.70 | |
| Self-Modification | 0.10 | |
| Dynamic Tool Use | 0.80 | |
| Persistent Memory | 0.50 | |
| Contextual Awareness | 0.80 | |
| Dynamic Identity | 0.30 | |
| 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 — Gander likely utilizes proprietary or fine-tuned LLMs for voice synthesis and text processing. Threats include prompt injection bypassing authentication or eligibility rules, and adversarial voice inputs manipulating the decisioning engine.
Not certain from the listing — Gander processes passenger PII, flight schedules, and claims data. Threats include data exfiltration of sensitive passenger details and poisoning of flight/disruption context data to trigger false compensation eligibility.
Gander orchestrates multi-step workflows for rebooking, accommodation, and reimbursement. Threats include tool misuse (unauthorized booking/payouts) and logic flaws in eligibility decisioning that can be exploited for financial gain.
Not certain from the listing — Likely hosted in cloud environments integrated with airline GDS (Global Distribution Systems) and telephony networks. Threats include unauthorized API access to GDS and compromise of telephony infrastructure leading to caller ID spoofing.
Not certain from the listing — Requires strict monitoring of automated payouts and voice call quality. Threats include blind spots in detecting fraudulent claims or prompt injection attacks during live calls.
Gander features 'PII-compliant customer authentication' and complies with passenger rights regulations (EU261, APPR, etc.). Threats include authentication bypass leading to unauthorized access to passenger accounts and regulatory non-compliance due to biased decisioning.
Not certain from the listing — May interact with external airline partner systems, hotel booking APIs, and payment gateways. Threats include cascading failures if external GDS or payment APIs are compromised or unavailable.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).