Innate — agentic threat model
Innate presents a high-risk profile due to its physical embodiment in home environments, where model manipulation or hijacking can translate directly into real-world property damage or physical safety hazards. The lack of visible security controls and its closed-source nature compound these risks.
OWASP AIVSS score rationale
| Autonomy of Action | 0.70 | |
| Goal-Driven Planning | 0.80 | |
| Self-Modification | 0.30 | |
| Dynamic Tool Use | 0.80 | |
| Persistent Memory | 0.90 | |
| Contextual Awareness | 0.90 | |
| Dynamic Identity | 0.10 | |
| Multi-Agent Interactions | 0.20 | |
| Non-Determinism | 0.80 | |
| Opacity & Reflexivity | 0.80 |
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 utilizes vision-language-action (VLA) or multimodal foundation models to translate user instructions into physical actions. Threats include physical adversarial attacks (e.g., visual patches that trick the robot's camera) and policy hijacking.
Not certain from the listing — relies on real-time spatial mapping, sensor telemetry, and user demonstration data for imitation learning. Threats include data poisoning during the 'teaching' phase and the unauthorized exfiltration of sensitive home layout or video data.
Not certain from the listing — orchestrates high-level goals into low-level motor control commands. Threats include insecure actuator tool integration, where a software-level prompt injection bypasses physical safety limits to cause rapid, destructive movements.
Not certain from the listing — likely deploys on edge hardware embedded in the robot, potentially communicating with cloud servers for heavy compute. Threats include physical device tampering, local privilege escalation, and insecure over-the-air (OTA) firmware updates.
Not certain from the listing — requires real-time physical safety guardrails (e.g., collision avoidance, emergency stop triggers). Threats include sensor spoofing causing blind spots, or software crashes that disable safety-critical monitoring loops.
Not certain from the listing — closed-source product with no publicly declared compliance frameworks (such as ISO/IEC 27001 or consumer robotics safety standards). Threats include lack of verifiable audit logs for physical actions and potential privacy violations regarding home surveillance.
Not certain from the listing — potential integration with broader smart home ecosystems (IoT). Threats include lateral movement where a compromised smart home device issues unauthorized commands to the physical robot.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).
These scores are auto-generated from public information (the agent's own listing, docs, and repository) using the canonical OWASP AIVSS formula and the MAESTRO framework — an estimate for guidance, not a penetration test, audit, or certification. See the scoring methodology. Are you the vendor? Factual corrections are free.