Ditto — agentic threat model
Ditto presents a high-consequence data privacy risk due to its handling of sensitive personal health information (PHI) combined with a low-to-moderate agentic profile focused on document parsing and sharing.
OWASP AIVSS score rationale
| Autonomy of Action | 0.20 | |
| Goal-Driven Planning | 0.20 | |
| Self-Modification | 0.00 | |
| Dynamic Tool Use | 0.30 | |
| Persistent Memory | 0.60 | |
| Contextual Awareness | 0.50 | |
| Dynamic Identity | 0.00 | |
| Multi-Agent Interactions | 0.00 | |
| Non-Determinism | 0.40 | |
| Opacity & Reflexivity | 0.40 |
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 — likely relies on third-party LLMs for document understanding. Key threats include prompt injection leading to medical misinterpretation or diagnostic hallucination.
Not certain from the listing — processes highly sensitive health documents and appointment data. Key threats include unauthorized data exfiltration of PHI and lack of secure local/cloud vector storage encryption.
Not certain from the listing — orchestration is required to parse documents and manage sharing workflows. Key threats include insecure tool integration with calendar APIs or sharing mechanisms.
Not certain from the listing — likely hosted as a mobile application with cloud backend support. Key threats include insecure local storage on mobile devices and API endpoint exposure.
Not certain from the listing — no mention of medical-grade guardrails or output verification. Key threats include silent failures in document parsing and lack of clinical accuracy monitoring.
Not certain from the listing — handling health data demands strict compliance (e.g., HIPAA, GDPR), but no specific compliance certifications or access controls are detailed.
Not certain from the listing — sharing features imply user-to-user or user-to-provider data flows. Key threats include unauthorized data sharing or interception by external entities.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).