Seedance — agentic threat model
Seedance is a generative video platform with low agentic risk, primarily acting as a direct text/image-to-video generator. Its main security risks lie in model abuse (e.g., deepfakes, NSFW generation) and resource exhaustion rather than autonomous execution or systemic lateral movement.
OWASP AIVSS score rationale
| Autonomy of Action | 0.10 | |
| Goal-Driven Planning | 0.10 | |
| Self-Modification | 0.00 | |
| Dynamic Tool Use | 0.10 | |
| Persistent Memory | 0.10 | |
| Contextual Awareness | 0.20 | |
| Dynamic Identity | 0.00 | |
| Multi-Agent Interactions | 0.00 | |
| Non-Determinism | 0.70 | |
| 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.
Utilizes text-to-video and image-to-video foundation models. Primary threats include adversarial prompt injections to bypass safety filters (generating deepfakes, copyright violations, or NSFW content) and model reprogramming.
Not certain from the listing — processes user-uploaded images and text prompts. If these assets are stored or used for downstream fine-tuning, there are risks of data leakage, intellectual property exposure, and training data poisoning.
Not certain from the listing — Seedance operates as a direct generation pipeline rather than an autonomous agent framework. Risks of tool misuse or complex planning vulnerabilities are minimal due to the lack of agentic orchestration.
Not certain from the listing — hosted as a freemium web service. High risk of GPU resource exhaustion attacks (denial of service) due to the computationally intensive nature of video generation.
Not certain from the listing — no details on content moderation guardrails or output monitoring. Gaps in real-time generation filtering could allow the platform to be abused for automated misinformation or harmful content generation.
Not certain from the listing — lacks explicit mention of enterprise security controls, user access management, or regulatory compliance (such as copyright protection or EU AI Act alignment for synthetic media).
Not certain from the listing — operates as a standalone horizontal tool with no described multi-agent interactions, marketplace integrations, or agent-to-agent trust boundaries.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).