Cybercrimes & the Underground Economy

Topics Actors · Underground Forums · Exploits-as-a-Service · Botnets · Spam · Scamming
01 //

Actors in the Underground

The cybercrime underground is populated by specialized actors who collaborate to compromise systems and monetize stolen data.

Primary Actors

Exploit Developers

Reverse-Engineers

Very smart people who reverse-engineer software. Develop and sell exploit packs and kits to other actors.

Botnet Masters

Zombie Controllers

Develop software and control vast numbers of zombie machines. Rent out their botnet to spammers, phishers, and other actors.

Spammers

Traffic Acquisition

Advertise links for other actors. Build, curate, buy, and sell email lists. Rent botnets to send bulk mail.

Supporting Actors

Phishers

Scam Sites

Setup scam sites to steal information. Work with spammers to spread phishing links to victims.

Counterfeiters

Fake Goods

Run websites selling fake goods (pharma, Rolex, etc.). Must be able to clear credit cards and ship products.

Bulletproof Hosting

Resilient Infrastructure

Offer dedicated servers in lawless parts of the Internet. Expensive but essential for long-running operations.

Carders, Cashiers & Mules

Monetization

Turn stolen bank accounts and credit cards into cash. Help launder money. Crucial link between theft and profit.

Crowdturfers

Fake Accounts & CAPTCHAs

Create, verify, and manage fake accounts via crowd-sourcing platforms. Solve CAPTCHAs for a fee.

02 //

Structure of the Underground

Interconnected Ecosystem

Bad actors form an interconnected ecosystem. Botnets are used for spam; spam facilitates phishing, counterfeit sales, and malware installation. Activities and infrastructures support each other.

The Value Chain

Underground Activities & Dependencies
Botnets

Core infrastructure for spam, DDoS, mining

Spam

Drives traffic to phishing, counterfeit, malware

Phishing

Steals credentials, bank logins, credit cards

Carders

Turn stolen accounts into cash

Pay-per-Install

Exploit-as-a-Service monetization

DDoS / Ransomware

Extortion using botnet power

03 //

Underground Forums

Large volumes of illicit goods and services are advertised on underground forums. Many operate in plain sight - a Google search away.

Roles & Dynamics

  • Buyers - seek stolen data, malware, hosting
  • Sellers - offer bank logins, hacked hosts, credit cards
  • Rippers - steal from naïve buyers or sell fraudulent goods
  • Administrators - verify trustworthy buyers, moderate deals
Law Enforcement Presence

Law enforcement often targets forums and IRC rooms. Some forums have been sting operations. New forums quickly fill the void when one is shut down.

Value for Researchers

White-Hat Intelligence

Black market forums are hugely valuable for security professionals. Allow researchers to observe trends, detect unfolding attacks, and understand the underworld.

/* Typical forum advertisements */ "I have BOA, Wells, Barclays bank logins..." "I have hacked hosts, mail lists, PHP mailer send to all inbox" "I need one MasterCard I give one Linux hacked root" "Verified PayPal accounts with good balance...can cash out PayPals"
04 //

Exploits-as-a-Service

Old Model
  • Compromise and monetization were coupled
  • Same criminal gang developed exploits, launched attacks, used hacked machines
vs
New Model - Decoupled
  • Exploit developers sell kits or packers
  • Other actors use kits to attack hosts (spam, compromised web servers)
  • Compromised hosts sold on black market - Pay-per-Install

Drive-by-Download & Exploit Kits

A website is compromised to embed malware in scripts. When a victim visits, the exploit kit installs malware. Blackhole, MPack, SpyEye, Zero Access, Rena FakeAV are examples.

Buy vs Rent

Two Styles

Buy: Miscreant buys an exploit kit and deploys it themselves.
Rent: Miscreant rents access to an exploit server that hosts the kit.

Traffic Acquisition

Victims Must Visit

Miscreants are responsible for acquiring traffic - directing victims to exploit kits via spam or phishing.

Traffic-PPI (Pay-per-Install)

Simplified Distribution

Traffic-PPI services bundle traffic acquisition and an exploit server. One service handles both: getting victims to the site and infecting them. Now the most popular way to distribute malware.

Traffic PPI Flow
Exploit Pack Developer
Payment
per install
Clients (Attackers)
Malware
Compromised Site / Redirect Chain
Victim

PPI Terminology

Term Definition
Doorway pages Webpage listing many keywords to increase search ranking; scripts redirect to attacker's page
Crypters Program that hides malicious code from anti-virus software
Blackhat SEO Manipulates search engines to increase traffic to attacker's site
Trojan Download Manager Software that lets attacker update or install malware on victim's computer

Deep Web vs Surface Web vs Dark Web

Surface Web

Readily available to the public, searchable with standard search engines. ~4% of WWW content.

Deep Web

Not indexed by standard search engines. Part of the Internet hidden from view. ~96% of WWW content.

Dark Web

Web content on darknets - overlay networks accessed via specific software (e.g., Tor).

05 //

From Malware to Botnets

Infected machines have valuable resources: spare CPU cycles, unique IP addresses, bandwidth. Botnets aggregate and control them via Command & Control (C&C) infrastructure.

Infected Machine Resources

CPU

Spare cycles for mining, cracking, computational tasks

IP & Bandwidth

Distributed IPs evade spam filters; bandwidth for DDoS

Scale

Swaths of bots rented to other actors for various purposes

Command & Control Architectures

IRC - Centralized

Single Point of Failure

Botmaster sends commands via IRC (e.g., sndspam: <subject>). Efficient but easy to locate and take down.

P2P - Structured DHT

Decentralized

Botmaster inserts commands into DHT. Bots get commands from peers. More robust but no direct synchronized control.

Fast Flux DNS

Evade Blocking

C&C website mapped to different IPs every ~10 seconds. Defeats IP-based blocking. But: ISPs can blacklist the rendezvous domain.

Random Domain Generation

Modern Approach

Bots generate many random-looking domains each day (shared algorithm + seed). Botmaster registers only a few. Hard to detect - domains are new, short-lived.

06 //

Spam & Conversion

>90%
Of Email is Spam
Hundreds of Billions
Spam Messages Per Day
Defining Characteristics of Spam

1. Inappropriate or irrelevant to the user.
2. Sent to a large number of recipients.

Spam Affiliate Marketing

Scammers set up websites (pharma, knockoffs, fake AV). Spammers sign up as affiliates, advertise the scams, and collect 30–50% commission on sales. Scammers deliver products and collect payments - many have customer service departments.

Spam Conversion Rate

The percentage of spam messages that result in a final sale. Key to understanding spam economics. Measurement: infiltrate spam generation/monetization and instrument the pipeline.

Storm Botnet Case Study

Researchers infiltrated Storm and measured filter effectiveness. On average only 0.014% of spam got through - 1 in 7,142. Gmail, Yahoo, Hotmail, Barracuda varied by campaign.

Storm: Pharmaceutical Revenue

Extrapolated from 1.5% of Botnet

28 purchases in 26 days, avg ~$100. Study controlled ~1.5% of workers.

$9,500/day
Extrapolated Revenue
$3.5M/year
Total Sales

40% cut for Storm operators via Glavmed affiliate program.

$1.7M/year
Storm Operators
Top Countries: Spam Cart Additions

United States, Canada, Philippines - visitors who added items to cart from spam-directed traffic.

07 //

Scamming Ain't Easy

Scamming requires a whole ecosystem: network infrastructure, payment processing, and bulletproof hosting. Example: setting up canadianpharma.com.

Infrastructure Choke-Points

Component Problem Solution
Domain name Legit registrars take down on complaints Shady registrars - charge more
DNS servers Obvious choke-point for law enforcement Bulletproof DNS - expensive
Web servers ISP/law enforcement can shut down Bulletproof hosting - expensive
Payment processing Most banks won't work with scammers Banks in lawless countries
Why Scammers Ship Products

Unhappy customers complain → processors shut down accounts → bank accounts seized. Scam sites almost always ship products to avoid chargebacks.

PharmaLeaks - Pharmacy Express

In 2012, GlavMed, SpamIt, RX-Promotion databases were breached and dumped. Researchers analyzed complete logs of sales, customers, affiliate relationships.

Findings
  • RX-Promotion and GlavMed ≈35% of affiliate scams
  • Repeat customers ≈33% of sales - counter to "pure scam" wisdom
  • ED drugs highest demand; pain meds, sleeping aids also popular
Economics
  • 30–40% commission to affiliates
  • ~10% of affiliates account for 75–90% of revenue
  • Net revenue modest - typically under 20% of sales
  • Payment processors = weak point in value chain
08 //

Summary & Key Takeaways

Underground Ecosystem

Actors

Exploit devs, botnet masters, spammers, phishers, counterfeiters, bulletproof hosting, carders, crowdturfers.

Exploits-as-a-Service

Decoupling: buy/rent kits, Traffic-PPI bundles traffic + exploit server.

Botnets

IRC (simple), P2P (robust), Fast Flux, Random domain generation for C&C.

Spam & Scamming

Spam Conversion

~0.014% get through filters. Storm: $3.5M/yr pharma, $1.7M to operators.

Scamming Infrastructure

Domain, DNS, web, payment - all choke-points. Bulletproof services expensive.

PharmaLeaks

Repeat customers, affiliate skew, payment processors as weak point.

The Takeaway

The cybercrime underground is a specialized, interconnected economy. Exploit developers, botnet masters, spammers, and phishers collaborate through forums and black markets. Monetization and compromise are decoupled via Pay-per-Install and Exploits-as-a-Service. Disrupting payment processors or top affiliates has disproportionate impact.

Further Reading

An Empirical Analysis of Spam Marketing Conversion - botnet infiltration to measure delivery and conversion.

Understanding the Business of Online Pharmaceutical Affiliate Programs - GlavMed/SpamIt data analysis.

Characterizing Spam-advertised Revenue - where spam-directed visitors add to cart.