---
name: Fraud
slug: fraud
type: social_exploit
status: running
version: 6.1.4
released: "prehistory"
maintainer: unknown (contributions from every era)
dependencies:
  - trust
  - information_asymmetry
  - plausible_deniability
  - a_victim_who_wants_to_believe
license: Proprietary (self-replicating)
tags:
  - crime
  - communication
  - economics
  - human_nature
  - runtime_error
---

Fraud

What it actually is

A performance in which one party rewrites the other's model of reality, then charges admission.

How it works

Fraud runs on a two-layer exploit stack.

Layer 1: The Gap A delta exists between what is true and what is believed. Fraud does not create this gap. It finds it, measures it, and invoices it.

Layer 2: The Consent Illusion The target consents to something. They just consent to a thing that does not exist. The signature is real. The contract is a prop. This is the technical elegance that separates fraud from theft. No broken windows. No forced entry. The door was opened from the inside.

Execution typically follows this loop:

while trust > 0:
    present false_reality as plausible
    collect value (money, data, labor, votes)
    if suspicion > threshold:
        reframe, delay, or disappear
    else:
        continue

Features

Scales horizontally. One fraudster, millions of targets. No marginal cost per lie.
Self-healing reputation layer. charisma patches most surface cracks before inspection.
Deeply cross-platform. Runs on romance, finance, politics, medicine, memoir, and the internet without recompilation.
Plausible deniability built in at the architecture level. Rarely ships a confession.

Known bugs

Occasionally the fraudster believes their own false_reality. This is classified as a feature by some downstream modules (see: delusion).
Evidence accumulates in proportion to ambition. Large frauds leave geological records.
Exit timing is notoriously miscalculated. Most crashes happen because the operator stayed one cycle too long.
Victims often debug the wrong layer, blaming themselves for the trust rather than the fraudster for the exploit.

Error codes

Code	Name	Description
`FR-001`	NullSubstance	Delivered product does not exist
`FR-002`	IdentityCollision	Operator is not who the interface claims
`FR-003`	PromissoryOverflow	More was promised than physics allows
`FR-404`	VictimNotFound	Target realized before value transfer completed
`FR-500`	OperatorExposure	The performance ended without the performer leaving

Edge cases

When fraud is performed at sufficient institutional scale, it is sometimes reclassified as policy.
When the fraudster dies before exposure, historians debate the correct label for decades.
When both parties are defrauding each other simultaneously, the system enters an undocumented state with no known resolution path.

FAQ

Is fraud always intentional? Legally, yes. Phenomenologically, it gets complicated around the edges where self-deception is a dependency.

What is the difference between fraud and marketing? Regulatory jurisdiction and font size.

Can fraud be a net positive? Edge cases exist. Most require a very long time horizon and considerable motivated reasoning to construct.

Changelog

v1.0 — Verbal promises, livestock-based losses
v3.2 — Paper instruments introduced. Surface area expanded dramatically.
v5.0 — Digital infrastructure. Latency dropped to near zero.
v6.1.4 — AI-assisted scaling. Patch notes pending legal review.

drafting spec…