Andrés Romero Castro

Two Ways to Think About Fraud

In fraud prevention, we often talk about behaviors, policies, tools, and outcomes. But beneath all of that, there are two foundational ways of thinking — two mental models that shape how decisions are made.

They’re not official. You won’t find them listed in a manual. But they guide choices every day, whether consciously or not. And once you see them, you can’t unsee them.

I call them the Loss-Based Model and the Intent-Based Model. Let’s define them.

The Loss-Based Model

This approach is built around the principle of financial harm. It starts with a practical, measurable question:

Did we lose money?

If yes → investigate, act, escalate.
If no → approve, ignore, or deprioritize.

This model is widely used in fraud operations, chargeback handling, and KPI-driven environments. It’s reliable, efficient, and easy to scale.

Focuses on:

  • Unauthorized payments
  • Account takeovers
  • Chargebacks and claims
  • Stolen credentials

Strengths:

  • Data-driven
  • Prioritizes immediate threats
  • Clear cost-benefit logic

Limits:

  • Can overlook long-term manipulation
  • Doesn’t capture intent or potential abuse
  • Reacts to what’s already happened

The Intent-Based Model

This approach shifts the focus from outcomes to motivation. Instead of asking whether money was lost, it asks:

What was the user trying to do?

It looks at:

  • Manipulative behavior
  • Pattern-based abuse
  • Platform misuse
  • Attempts to deceive (even without financial impact)

Strengths:

  • Captures emerging fraud trends
  • Prevents reputational or systemic risks
  • Helps shape ethical and resilient systems

Limits:

  • Slower to act
  • Requires judgment
  • Harder to automate or scale

But it allows us to see beyond the surface — and understand when a “legitimate transaction” might not be so innocent after all.

What about potential losses?

One of the biggest challenges with the loss-based model is that it often leads us to focus on the wrong losses — the ones that are easy to count, but not the ones that really matter.

We end up preventing minor or repeated losses, the kind that look good on a dashboard. But the big losses — the painful ones — usually come from patterns we didn’t catch in time.

And that’s where the intent-based model shines. By understanding behavior early, even before it becomes “fraud,” we have a chance to stop something before it scales — before it costs real trust, money, or reputation.

So… what do we do?

This kind of reflection raises important questions:

  • Should we act when we see something dishonest, even if we’re not the ones being harmed?
  • Should we allow people to use a system to mislead others?
  • What kind of responsibility do we have when we see something that’s not “illegal,” but clearly isn’t right either?

And maybe more importantly:

  • Do we only react to what hurts now, or do we act to prevent what might hurt later?

There’s no perfect answer. But the conversation matters.

Not every act of deception leaves a loss. But that doesn’t mean it leaves no trace.

—

This reflection introduces the “Loss-Based vs. Intent-Based” model as a personal framework for thinking about fraud. While similar distinctions exist across the industry, these terms are used here to describe common but often unnamed patterns in fraud decision-making.

In