Dialoft AI — agentic threat model
Dialoft AI presents a moderate-to-high agentic risk due to its integration with telephony systems and CRMs, enabling automated outbound calling and data modification. A compromise could lead to automated vishing campaigns, PII exfiltration, and unauthorized CRM tampering.
OWASP AIVSS score rationale
| Autonomy of Action | 0.80 | |
| Goal-Driven Planning | 0.60 | |
| Self-Modification | 0.10 | |
| Dynamic Tool Use | 0.70 | |
| Persistent Memory | 0.50 | |
| Contextual Awareness | 0.60 | |
| Dynamic Identity | 0.20 | |
| Multi-Agent Interactions | 0.10 | |
| 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 — likely utilizes third-party LLMs and TTS/STT engines. Primary threats include voice-based prompt injection (over-the-air injection by callers) and model hallucinations leading to incorrect commitments during live calls.
Not certain from the listing — processes uploaded contact lists and CRM data. Threats include unauthorized access to or exfiltration of customer PII, and data poisoning via malicious contact list uploads.
Orchestrates voice interactions, CRM updates, and appointment scheduling. Threats include insecure tool integration with CRMs, where a caller manipulates the agent into executing unauthorized database writes or scheduling conflicts.
Not certain from the listing — operates as a cloud-hosted SaaS platform with telephony infrastructure. Threats include SIP/telephony trunk abuse, toll fraud, and unauthorized access to API keys used for CRM integrations.
Provides call summaries, sentiment analysis, and analytics. Threats include callers successfully social-engineering the agent while the agent's summary generation fails to flag the anomalous behavior, creating an observability blind spot.
Not certain from the listing — handles PII and automated outbound dialing, which subjects it to strict regulatory frameworks like TCPA, GDPR, and CCPA. No specific security certifications (e.g., SOC2) are detailed in the listing.
Not certain from the listing — primarily operates in a single-agent-to-human paradigm. The ecosystem risk is concentrated on the trust boundary between the voice agent and the connected CRM/scheduling APIs.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).