Agentforce — agentic threat model
Agentforce presents a high-impact risk profile due to its deep integration with enterprise CRM data and its autonomous capabilities in customer-facing sales and support roles. While Salesforce's platform security and human-in-the-loop blending provide mitigations, unauthorized tool execution or data exfiltration remain critical concerns.
OWASP AIVSS score rationale
| Autonomy of Action | 0.80 | |
| Goal-Driven Planning | 0.70 | |
| Self-Modification | 0.20 | |
| Dynamic Tool Use | 0.80 | |
| Persistent Memory | 0.80 | |
| Contextual Awareness | 0.90 | |
| Dynamic Identity | 0.50 | |
| Multi-Agent Interactions | 0.60 | |
| Non-Determinism | 0.70 | |
| Opacity & Reflexivity | 0.60 |
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 powering the Einstein 1 Platform are not detailed, but they likely face standard LLM threats such as prompt injection, jailbreaking, and output hallucination during customer interactions.
Leverages real-time data harmonization and Salesforce Data Cloud integration. Threats include unauthorized data access, RAG-based data leakage of sensitive customer PII, and potential data poisoning of the CRM knowledge base.
Utilizes a low-code/no-code development environment for orchestration. Risks involve insecure tool integration, unauthorized cross-platform automation, and logic flaws in autonomous SDR or Service Agent workflows.
Hosted within Salesforce's enterprise cloud infrastructure (Einstein 1). Primary threats include API vulnerabilities, tenant isolation failures, and unauthorized access to integration secrets.
Features a blend of digital and human-assisted support, implying some level of human-in-the-loop monitoring. Gaps may exist in real-time drift detection or logging of autonomous negotiation simulations.
As a closed-source, paid enterprise platform, it inherits Salesforce's compliance frameworks, but risks remain regarding dynamic consent, data residency, and auditability of autonomous agent decisions.
Features multiple specialized agents (SDR, Coach, Service). Threats include cascading failures across automated workflows, trust abuse between agents, and unauthorized horizontal escalation within the CRM ecosystem.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).