Field Reference · AI-Era Fraud

Fraud in the
AI Era.

A complete field reference. Threat taxonomy, defense architecture, and primary source ledger. Every claim traces to a government or institutional source. Updated continuously.

Updated May 2026
65+ Primary Sources
8 Threat Categories
6 Defense Protocols
Master Index

Navigate the reference.

Two indexes. Locate your threat. Locate your defense. Jump directly to the relevant section.

Architecture of Fraud

Three principles. Unchanged since the first con.

AI changed scale and realism. The underlying mechanism is identical to every fraud operation ever run.

01 · Psychology
Scams are not technical. They are psychological.
Urgency. Authority. Fear. Greed. Four triggers. Every scam in history runs on one or more of them. The technology is the delivery mechanism. The exploit is always human.
02 · What AI Changed
Scale and realism. Nothing else.
Perfect grammar. Voice cloning from seconds of audio. Real-time deepfake video. Personalized attacks from harvested data. The tell is gone. The mechanism is the same.
03 · The Equation
Believability × Urgency × Emotional Trigger.
Maximize any one variable and compliance follows. The defense is time. Every second you slow down degrades all three variables simultaneously. Calmness is the architecture.
Believability  ×  Urgency  ×  Emotional Trigger  =  Compliance
Your calmness sets every variable to zero.
The permanent defense.

You can hang up. Right now. In the middle of any conversation. With anyone. You do not owe a stranger on the phone your time, your attention, or your information. That feeling of urgency on a call or in a message — that is the attack. Your calmness is the defense.

Threat Taxonomy

Eight documented threats. Primary sources attached.

Every entry: mechanism, scale, documented case, defense protocol, and primary source citation.

Voice Cloning — AI Synthesis of Family Members
Mechanism: social media audio extraction → model training → real-time synthesis
Active

A scammer extracts thirty seconds of audio from a social media video. They feed it into a voice cloning model. Within minutes they have a synthetic voice indistinguishable from your family member — including the way they cry, the cadence of panic. They call you. You hear your child's voice saying they've been arrested. The voice is real. The emergency is fabricated.

Voice cloning is also deployed against bank voice-authentication systems. A synthesized version of your voice passes the biometric gate. The bank hears you. You gave them no such permission.

30 sec Audio required to clone a voice
1 in 4 Americans have received an AI voice call
442% Surge in vishing attacks · 2025
Defense Protocol

Establish a family codeword now. A phrase known only to immediate family. Any emergency call must use it before any action is taken. Hang up on any distress call and call the person directly at a saved number. Set social media to private. Restrict who can see videos containing your voice.

↗ FTC · Preventing Harms of AI-Enabled Voice Cloning ↗ FTC Alert · Fighting Back Against Voice Cloning ↗ FBI IC3 · AI-Generated Content in Fraud Operations
Deepfake Image & Video Extortion
Mechanism: public photo scraping → AI generation → coercion for payment
Active

Scammers scrape public photos from social media and feed them into AI generation tools. They produce fabricated images or videos of you in compromising situations that never occurred. They contact you demanding payment to suppress the material. This is happening to ordinary people — not public figures.

The more publicly accessible photos you have online, the larger the scammer's material library. A private profile is a meaningful structural defense.

700% Surge in deepfake video scams · 2025
Defense Protocol
  • Set all social media profiles to private immediately.
  • Audit what photos of you exist publicly. Remove what you can.
  • If targeted — do not pay. Payment confirms viability. Demands compound.
  • Document everything. Report to FBI IC3 at ic3.gov.
↗ FBI IC3 · Internet Crime Complaint Center ↗ FinCEN · Deepfakes in Financial Fraud · Nov 2024 ↗ Security.org · Deepfake Statistics 2025
SIM Swapping — Your Phone Number Stolen Without Touching Your Device
Mechanism: carrier social engineering → number transfer → 2FA interception
Active

A scammer calls your mobile carrier with your name, address, and the last four digits of your Social Security number — likely purchased from a data breach. They claim to have lost their phone and need a replacement SIM. Your number transfers to a SIM they control. Every call, every text, every two-factor authentication code now routes to them. Every bank alert. Every password reset link.

In 2025 a California arbitrator ordered T-Mobile to pay $33 million after a SIM swap attack enabled the theft of $38 million in cryptocurrency from a customer who had extra security measures on their account. SIM swapping defeats SMS-based two-factor authentication completely.

$50M FBI reported SIM swap losses · 2023
$33M T-Mobile judgment · SIM swap negligence
1,055% Surge in UK SIM swap cases · 2024
Defense Protocol
  • Call your carrier today. Request SIM lock, port freeze, and supervisor-required account PIN.
  • Replace SMS-based 2FA with an authenticator app — Google Authenticator, Authy, Microsoft Authenticator.
  • For high-value accounts use a hardware security key. A YubiKey cannot be defeated by SIM swapping — it never touches the cellular network.
↗ DeepStrike · SIM Swap Statistics 2025 ↗ OWASP · SIM Swapping Prevention Guidelines ↗ CTIA · Wireless Industry SIM Swap Protection
AI Phishing — Indistinguishable from Legitimate Communication
Mechanism: LLM generation → personalized targeting → credential/payment harvest
Active

For decades the tell of a phishing email was broken grammar and generic greetings. That tell is gone. Large language models produce flawless, contextually appropriate, professionally crafted messages in any language at zero cost and at scale. 82.6% of phishing emails now contain AI-generated content. The mechanism is unchanged. The quality is indistinguishable from legitimate communication.

82.6% Of phishing emails contain AI-generated content
$21B Total internet crime losses · 2025
Defense Protocol

If you did not initiate the contact, do not click any link in any message. Close it. Navigate directly to the institution's website through your browser or call the number on the back of your card. The link in the message is the attack. The message is the delivery mechanism.

↗ FBI IC3 PSA · AI-Generated Fraud Content 2024 ↗ FBI · Cryptocurrency and AI Scams · 2025 Report ↗ Proofpoint · Email Attacks Driving Record Losses 2024 ↗ Vectra AI · AI Scams — Enterprise Detection Analysis
Deepfake Video Calls — Real-Time Face and Voice Synthesis
Mechanism: identity model training → real-time video overlay → executive/family impersonation
Emerging

In 2024 a finance worker in Hong Kong was tricked into transferring $25 million after a deepfake video call impersonated the company's CFO and other colleagues. The worker attended what appeared to be a legitimate multi-person video conference. Every face on the call was synthetic. Real-time deepfake video is no longer theoretical — it is being deployed in targeted attacks against financial decision-makers.

$25M Lost in single deepfake video call · Hong Kong 2024
Defense Protocol

For any high-stakes decision made over video call — financial transfers, access grants, sensitive information — establish a separate verification step through a known channel before acting. Any caller who resists verification for a high-stakes request is the verification you needed.

↗ FinCEN · Alert on Deepfakes in Financial Fraud ↗ Keepnet · Deepfake Statistics Through 2026
Pig Butchering — Crypto Investment and Romance Scams
Mechanism: manufactured trust over weeks/months → fake investment platform → exit with funds
Active

Someone contacts you online. They are warm, attentive, patient. Over weeks or months they build genuine trust. Then they introduce an investment opportunity — typically cryptocurrency on a platform you have never heard of. They show you impressive returns. They encourage you to invest more. When you attempt to withdraw, the platform invents reasons you cannot access your funds. The relationship, the warmth, the patience — all of it was manufactured. Pig butchering revenue grew 40% in 2024. Total crypto fraud hit a record $9.9 billion that year.

$9.9B Crypto scam losses · 2024 record
40% Pig butchering revenue growth · 2024
Defense Protocol

Any investment opportunity introduced by someone you met online is a scam until independently verified through a licensed financial institution. Never send money to someone you have only met online regardless of how long the relationship has developed. Tell a family member about any new online relationship that moves toward financial discussion.

↗ Chainalysis · Pig Butchering Revenue 2024 ↗ CNBC · Crypto Scams Hit Record $9.9B in 2024 ↗ arXiv · Lifecycle of Pig Butchering Scams 2025
Government Impersonation — IRS, SSA, Medicare, Law Enforcement
Mechanism: authority claim → urgency trigger → payment demand via untraceable method
Active

The IRS does not call. The Social Security Administration does not call to tell you your number has been suspended. Medicare does not call to verify benefits. Law enforcement does not call to tell you there is a warrant that can be resolved with a gift card payment. If any caller claims to be from a government agency and demands immediate payment in any form — it is a scam. The real agency will send a letter. Caller ID is fakeable. The number displayed proves nothing.

Defense Protocol

Hang up. Find the agency's real phone number through a web search or a physical document you already have — never call back the number that called you. Every government agency with legitimate business sends written notice first. Payment demand by phone from a government agency is the defining signature of a scam.

↗ FTC · How to Avoid Imposter Scams ↗ FTC Alert · Impersonation Scams 2025 ↗ IRS · Recognize Tax Scams and Fraud
Tech Support Scams — Remote Access and Credential Theft
Mechanism: fake alert → phone call → remote access granted → full credential exposure
Active

A pop-up appears on your screen warning your computer has been infected. It displays a phone number. The person on the line is professional and helpful. They ask for remote access to your computer. Once they have it they can see everything — saved passwords, banking information, stored files, session cookies for every account you are logged into. Microsoft, Apple, and Google do not send unsolicited pop-up warnings with phone numbers to call.

Defense Protocol
  • Close the pop-up. If you cannot close it, restart your computer.
  • Do not call the number displayed.
  • Do not allow anyone who contacted you unsolicited to remotely access your device for any reason.
  • Legitimate tech support does not initiate contact with you.
↗ FTC · Master Scam Reference Hub
Defense Architecture

Six protocols. Implementable today.

Organized by the principle that governs all of them: compartmentalization. The goal is to limit the blast radius of any single compromise.

01 · Compartmentalize Your Digital Identity

Banking credentials, personal email, retail accounts, social media, and work accounts must exist in separate compartments that share no passwords, email addresses, or phone numbers.

  • Create a retail-only email address. Use it exclusively for store loyalty programs, online shopping, and any service that doesn't involve banking. Never attach financial accounts to it.
  • Use a separate phone number for retail signups. Google Voice provides a free number. Your real number stays protected.
  • Never reuse passwords. A password manager — Bitwarden is free and open source — generates and stores unique passwords for every account. This single change eliminates the most common account takeover vector.
  • Prepaid debit card for in-store purchases. Damage from any single compromise is capped at the card balance.
Free Implementable today Defeats 6 threat categories
02 · The 2FA Hierarchy — Not All Two-Factor Authentication Is Equal
  • SMS codes — weakest. Defeated entirely by SIM swapping. Replace wherever possible.
  • Authenticator apps — strong. Google Authenticator, Authy, Microsoft Authenticator generate codes locally on your device. A SIM swap does not affect them. Use for email, banking, and every account that supports them.
  • Hardware security keys — strongest. A YubiKey or FIDO2-compliant device is a physical key. It cannot be intercepted remotely. It cannot be phished. It cannot be defeated by SIM swapping. Use for banking, primary email, and your password manager.
SMS → Authenticator → Hardware key YubiKey: ~$50 Authenticator apps: free
↗ NIST · MFA Guidance ↗ NIST SP 800-63B · Official Digital Identity Guidelines
03 · Carrier Lock Protocol

Call your mobile carrier today. Request all three of the following. Document the representative's name and the date.

  • SIM lock — prevents your SIM from being swapped without in-person verification.
  • Port freeze — prevents your number from being transferred to another carrier.
  • Account-level PIN — requires supervisor approval before any account change can occur.
  • Ask specifically whether your carrier offers a verbal password requirement. If they do — set it.
Free One phone call Defeats SIM swapping entirely
↗ Verizon · Enable SIM Protection
04 · Freeze Your Credit at All Three Bureaus

A credit freeze prevents any new credit from being opened in your name without your explicit unfreeze. It is free. It does not affect your existing accounts or your credit score. It is the single most effective defense against identity theft. A freeze at one bureau does not cover the others — all three are required.

  • Equifaxequifax.com · 800-685-1111
  • Experianexperian.com · 800-311-4769
  • TransUniontransunion.com · 800-888-4213
  • Pull one free credit report every four months — staggered across bureaus — for continuous visibility. Official source only: annualcreditreport.com or 877-322-8228.
Free Does not affect credit score Reversible when needed
↗ FTC · Credit Freezes and Fraud Alerts ↗ USA.gov · Credit Freeze Instructions
05 · Payment Hierarchy
  • Credit cards — highest protection. Federal law limits your liability to $50 for unauthorized charges. Most issuers eliminate liability entirely if reported promptly. Use for online purchases wherever possible.
  • Digital wallets — strong. Apple Pay and Google Pay use tokenization — your actual card number is never transmitted. Use tap-to-pay wherever available.
  • Gift cards, wire transfers, cryptocurrency, Zelle, Venmo — unrecoverable. If anyone insists you pay by any of these methods — stop. This is the defining signature of a scam. Once money leaves through these channels it does not return.
Gift card demand = scam. Always.
06 · Family Codeword — The Only Defense Against Voice Cloning

Agree on a word or phrase known only to your immediate family. Any emergency call requesting money or action must include the codeword before any response is given. A voice clone cannot guess a word it has never heard. This is the primary structural defense against the grandparent scam and every variant of it.

  • Choose the word now. Tell immediate family only. Write it down somewhere secure.
  • Hang up on any distress call — however convincing — and call the person at a number you already have saved before taking any action.
  • The caller who resists the codeword protocol is the confirmation the call is fraudulent.
Free Takes five minutes Defeats voice cloning entirely
Response Protocol

If it happens. Move fast.

Every hour that passes is an hour the attacker has to use what they took. Speed is the primary variable.

01
Document everything immediately

The phone number. What was said. What information you provided. What payment you made. What accounts were mentioned. Do this while it is fresh. This documentation is essential for every subsequent step including law enforcement reporting.

02
Change passwords on every affected account

Every account where you used the same password as a compromised account. Every account linked to a compromised email. Change to unique passwords. Enable the strongest available 2FA on each. Act before the attacker does — do not wait for signs of compromise.

03
Contact your financial institutions

Call your bank and credit card companies using the number on the back of your card. Report what happened. Request account flags for suspicious activity. Credit card payments may be reversible. Wire transfers, gift cards, cryptocurrency, and Zelle payments generally are not — report them anyway. Documentation matters for law enforcement.

04
Freeze your credit at all three bureaus

Equifax · Experian · TransUnion. A freeze at one does not cover the others. Free. Reversible. Do it before any new credit can be opened in your name.

05
Report it to federal authorities

File with the FTC at ReportFraud.ftc.gov. File with the FBI's Internet Crime Complaint Center at ic3.gov. Report to local law enforcement as well. These reports contribute to pattern analysis that identifies and prosecutes criminal networks. Your report matters even if the individual loss cannot be recovered.

06
Tell someone

The shame of being scammed is one of the scammer's most powerful tools. It keeps victims silent. Silent victims do not report. Silent victims do not warn their communities. The scammer counts on this silence to continue operating. The embarrassment belongs to the criminal. Never to the person who was targeted.

Primary Source Ledger

Resources & Provenance

This document cites no secondary sources where a primary exists. Every institution, report, and dataset below is the origin record — not an article about it.

Government & Law Enforcement
FTC — Master Scam Reference Hub
Continuously updated. Master list of scam types, active alerts, prevention strategies, and psychological manipulation analysis.
Federal Trade Commission · consumer.ftc.gov/scams
ftc.govGov · Primary
FTC — How to Avoid Imposter Scams
Definitive FTC guidance: government agencies never demand immediate payment or threaten arrest by phone.
Federal Trade Commission · 2025
ftc.govGov · Primary
FTC Alert — Impersonation Scams 2025
Caller ID is fakeable. Unsolicited contact from any authority is a red flag regardless of displayed number.
Federal Trade Commission · April 2025
ftc.govGov · Primary
FTC Alert — Voice Cloning
AI can mimic voices convincingly from minimal audio. Family verification protocols are the structural defense.
Federal Trade Commission · April 2024
ftc.govGov · Primary
FTC — Preventing Harms of AI-Enabled Voice Cloning
FTC policy analysis on the systemic risks of voice cloning technology deployed at scale against consumers.
Federal Trade Commission · November 2023
ftc.govGov · Primary
FTC — Proposed Rule: AI Impersonation of Individuals
FTC regulatory action to address AI-generated impersonation fraud at the statutory level.
Federal Trade Commission · February 2024
ftc.govGov · Primary
FTC — Report Fraud
Reports shared with over 2,000 law enforcement agencies. Primary reporting mechanism for consumer fraud.
Federal Trade Commission · reportfraud.ftc.gov
ftc.govGov · Report
FBI — Internet Crime Complaint Center (IC3)
Primary federal reporting and tracking system for internet crime. Source for annual cybercrime loss statistics.
Federal Bureau of Investigation · ic3.gov
ic3.govGov · Primary
FBI IC3 — 2024 Annual Report
$16.6 billion in reported cybercrime losses in 2024. Comprehensive breakdown by fraud type, victim demographics, and state.
Federal Bureau of Investigation · 2024
ic3.govGov · Report
FBI IC3 PSA — AI-Generated Content in Fraud
FBI public service announcement documenting AI-generated text, images, and audio being deployed in phishing, romance, and investment scams.
Federal Bureau of Investigation · December 2024
ic3.govGov · PSA
FBI — Cryptocurrency and AI Scams · 2025 Report
Approximately $21 billion lost to internet crime in 2025. AI and cryptocurrency fraud constituting the fastest-growing loss categories.
Federal Bureau of Investigation · 2025
fbi.govGov · Primary
FinCEN — Alert on Deepfakes in Financial Fraud
Financial Crimes Enforcement Network alert on deepfake technology being deployed against financial institutions and their customers.
U.S. Department of the Treasury · November 2024
fincen.govGov · Alert
IRS — Recognize Tax Scams and Fraud
IRS primary guidance on how the agency does and does not contact taxpayers. The IRS initiates contact by mail, not phone.
Internal Revenue Service · irs.gov
irs.govGov · Primary
CFPB — Identity Theft Recovery Guide
Consumer Financial Protection Bureau official identity theft response protocol including credit reporting and dispute procedures.
Consumer Financial Protection Bureau · consumerfinance.gov
cfpb.govGov · Primary
FTC — Credit Freezes and Fraud Alerts
Definitive FTC guidance on credit freeze procedures, fraud alert types, and the distinction between them.
Federal Trade Commission · consumer.ftc.gov
ftc.govGov · Primary
USA.gov — Credit Freeze Instructions
Official U.S. government portal with step-by-step credit freeze instructions for all three bureaus.
U.S. Government · usa.gov
usa.govGov · Primary
NIST — Multi-Factor Authentication Guidance
National Institute of Standards and Technology MFA implementation guidance including authentication tier definitions.
National Institute of Standards and Technology · nist.gov
nist.govGov · Primary
NIST SP 800-63B — Digital Identity Guidelines
The authoritative federal standard for digital authentication assurance levels. Source for authentication tier hierarchy used in this document.
National Institute of Standards and Technology · 2017, updated
nist.govGov · Standard
Fraud Prevention Field Reference is a publication of the Global Data Registry  —  open provenance infrastructure for the machine-readable web.
View the Registry →