Why Scams Got 10х Worse In 2 Years

What A Scam Looked Like In 2023

You probably remember the classic phishing email:

Subject: URGENT!!! Your acount has been sosspended!!

Dear User,

Your acount will be suspended within 24 hours unless you very your detals at the following link.

[click here urgent]

Best regards, Customer service department

You spotted it in three seconds. The typos, the generic greeting, the suspicious link, the weird capitalization. Even if you didn't know the word "phishing," you knew this email was wrong.

That world is gone.

What A Scam Looks Like In 2026

Same scam, modern execution:

From: Customer Service service@bank-of-spain.es

Subject: Quick check on your recent transaction

Hi Maria,

We noticed a card transaction at Starbucks Madrid Centro on 14:32 today for €4.85. Just confirming this was you?

If yes, no action needed — your card is active and protected. If no, please review the transaction here and we'll block the card immediately: [verify-transaction.bank-of-spain.es]

Best regards, Maria Lopez, Fraud Prevention Team Bank of Spain · Madrid Office

It's grammatically perfect. It uses your name. It references a transaction at the right Starbucks, at a plausible time. The "from" domain looks real. The link looks real. The signature has a name and a department.

There is nothing in this email that a human could spot as wrong on first read.

This is the new baseline. And it took the scammer about six seconds to generate.

What Changed — Three Things, All At Once

1. Generative AI dropped the cost of perfect scam content to zero.

The same large language models that help you write emails now write scam emails at industrial scale. No more bad grammar. No more typos. No more generic "Dear User" openings — AI personalizes every email with data scraped from your LinkedIn, your public social media, breach databases, and data broker sites.

The statistics, all from the last 18 months:

• AI-generated phishing now accounts for over 82% of all phishing emails;
• Phishing volume rose 1,200% in 2024 alone (FBI IC3 report);
• AI-enabled scams rose 456% in Q1 2025 compared to a year earlier (Sift).

The traditional advice to "look for bad grammar or weird wording" no longer works. There isn't any.

2. Voice cloning got real, fast, and cheap.

Two years ago, voice cloning required hours of source audio and produced something that sounded mostly like the target. Today, modern tools need about 3 seconds of source audio to produce an 85%-accurate clone. Anyone with a public TikTok, an Instagram Reel, or a voicemail greeting is now a viable target.

Voice clones drive an explosive new category of scam: vishing (voice phishing). The phone call from "your daughter" who's been in an accident. The call from "your boss" needing an urgent wire transfer. The Q1 2025 surge in deepfake vishing in the US was 1,600% year over year.

3. Scam operations went industrial.

The mental image of "a lone hacker in a basement" is wrong. Modern scam operations are run as professional businesses:

• Multi-story call centers in Cambodia, Myanmar, and Laos, with thousands of staff (many of whom are themselves trafficking victims forced to run scams);
• Multilingual support — same scam delivered fluently in 20+ languages;
• AI agents that maintain "personal" relationships with thousands of victims simultaneously, each conversation feeling intimate and unique;
• Mirror dashboards — pixel-perfect fake trading platforms, sometimes available in real app stores, that display fabricated profits to keep victims investing more.

The 2026 evolution of pig butchering scams (covered in Chapter 4) is the most extreme example — but the same tooling powers every other scam category.

What Did NOT Change — The Three Constants

Here's the part nobody tells you: the underlying psychology of every scam is the same as it's been for a thousand years. Better paint job. Same mechanics. Every scam, AI-powered or not, needs three things:

They want you in a hurry. Real situations almost never require an action in the next 60 minutes. Urgency is the scammer's friend because it bypasses your judgment.

They want you afraid. Fear of loss (your account, your money), fear of harm (a loved one), fear of consequences (legal action, exposure). Scared brains make worse decisions than calm brains. The scammer needs you scared.

They want you alone. Don't call your spouse. Don't tell your bank. Don't post about it. This is confidential. Why? Because outside perspective is the single most effective scam-killer in the world. A friend with no emotional stake will spot the scam in 10 seconds.

If you remember these three constants and nothing else from this section, you'll catch 90% of scams just by noticing the pattern: am I being rushed, scared, and isolated? If yes, stop everything.

What This Section Will Teach You

Five chapters, each one a specific modern scam type:

• Chapter 2 — Phishing emails. How to read one like a pro in 30 seconds;
• Chapter 3 — The deepfake "family emergency" voice call. Why a single secret word stops it cold;
• Chapter 4 — Romance scams and pig butchering. The slow, months-long manipulation;
• Chapter 5 — QR-code traps and clone shopping sites. The new physical-world attacks;
• Chapter 6 — The universal 5-question filter that catches almost everything, including scams that don't fit any neat category.

After this section, you'll handle the modern scam landscape with the same instinctive ease you handle "Nigerian prince" emails today.

Let's go.