The Human Layer of Fraud, Scams, and Cybercrime

An Executive Whitepaper to the HVE™ Framework

How behavioral science, psychology, criminology, and agentic AI create a new operating model for fraud and scam prevention, intervention, and resolution.

Fraud, scams, and cybercrime 
are human from the first signal.

Existing systems analyze events and artifacts such as transactions, devices, IPs, logins, claims, and tickets, but often miss the human exploitation pattern behind them.
The HVE™ model gives Charm AI Agents a Dual Brain: the investigative intelligence to analyze the evidence and the behavioral intelligence to understand context, deception, manipulation, intent, and emotion.

From Human Understanding to Action

HVE™ turns behavioral science, psychology, and criminology into a structured reasoning layer for AI agents.

The attack

The manipulation pattern being used.

The human context

The vulnerability or state being exploited.

The severity

The urgency and potential impact of the situation.

The action

The intervention or response required next.

Explore How Fraud Expertise and Behavioral Science Work Together

Traditional systems analyze alerts. The HVE™ model helps AI Agents to understand the human context.

Explore The HVE™ Framework

Download the whitepaper to discover how the HVE™ model can support scam and fraud prevention, intervention, resolution and disruption.

Contribute to the HVE™ Framework

No single institution has a complete view of human vulnerabilities and the techniques fraudsters use to exploit them. Charm is opening the HVE™ Framework to contributions from financial institutions, fraud teams, researchers, behavioral experts, victim-support organizations, and others working to prevent human-centric attacks.
Contributions may include anonymized cases, exploitation patterns, intervention outcomes, operational insights, and relevant research. Together, these contributions can strengthen the industry’s shared understanding of how people are targeted, manipulated, and protected, and help improve the interventions used to prevent harm.
Contribute to the HVE™ Framework and help build the human vulnerability layer together:
Build the Human Vulnerability Layer Together
hve@charmsecurity.com