How does Vyvern ensure personas and workflows are safe?
Vyvern uses both automated and manual review systems to help ensure personas and workflows remain realistic, appropriate, and aligned with organizational restrictions.
This includes:
AI-based safety and realism review
Administrative approval systems
Compliance validation
Organizational workflow restrictions
These systems work together to help reduce unrealistic, unsafe, or inappropriate workflow behavior.
AI-Based Review
Before workflows are executed, Vyvern’s AI evaluates personas and workflows for:
Realism and plausibility
Appropriate communication behavior
Workflow consistency
Target alignment
Organizational restrictions and context
The AI also helps identify:
Unrealistic persona behavior
Poorly aligned workflow objectives
Unsafe workflow patterns
Invalid targeting scenarios
This helps improve workflow quality while reducing inaccurate or unrealistic testing conditions.
Administrative Review
Organizations and MSP administrators also maintain direct control over workflow approval.
Depending on organizational settings:
Workflows may require manual approval
Compliance review may be enforced
Autopilot restrictions may apply
High-priority workflows may require additional validation
This allows organizations to maintain oversight over workflow execution and testing behavior.
Initial Workflow Review for New Organizations
For newly onboarded organizations, Vyvern manually reviews the initial workflows generated by the platform.
This review process helps ensure:
The AI correctly understands the organization
Workflow targets are appropriate
Personas align with organizational context
Workflow behavior remains safe and realistic
Additional review during onboarding helps improve long-term workflow quality and automation reliability.
Compliance and Organizational Restrictions
Organizations may configure restrictions around:
Workflow categories
Approval thresholds
Compliance requirements
Autopilot behavior
Target validation rules
These restrictions help organizations maintain control over how workflows and personas are generated and approved.
Why realism matters
Vyvern is designed to simulate realistic social engineering and workflow behavior while maintaining organizational safety and oversight.
Poorly designed or unrealistic personas can:
Reduce testing quality
Create unrealistic employee behavior
Lower trust realism
Produce inaccurate security assessments
By combining AI review with administrator oversight, Vyvern helps organizations maintain more accurate and controlled testing environments.