Build Your Own Botnet (CRAX - Pro)
Build Your Own Botnet (CRAX - Pro)
Build Your Own Botnet (CRAX - Pro)
Have a plan to steal millions from banks and their customers but can’t
write a line of code? Want to get rich quick off advertising click fraud but
“quick” doesn’t include time to learn how to do it? No problem.
Everything you need to start a life of cybercrime is just a few clicks (and
many more dollars) away.
In the process, these big botnet platforms have created a whole ecosystem
of software and services in an underground market catering to criminals
without the skills to build it themselves. As a result, the tools and
techniques used by last years’ big professional bank fraud operations, such
as the “Operation High Roller” botnet that netted over $70 million last
summer, are available off-the-shelf on the Internet. They even come with
full technical support to help you get up and running.
The customers of these services often plan more for the short term than the
long game played by the big cyber-crime rings. They have very different
goals. Botnet infrastructures can be applied in lots of ways for different
sorts of profit—cash, information, or political gain. There are many ways
to make money off botnets beyond outright theft, such as using them to
steal advertising clicks, generate spam e-mails for a paying client, or
renting out bots for denial-of-service attacks. And the same basic
principles used to distribute botnets have been creeping up in more
targeted attacks to steal intellectual property or to spread the malware used
in the recent “wiper” attack on South Korean banks and broadcasters.
So how easy is it to get into the botnet business? Well, Ars decided to find
out. Given the surprising availability of botnet building blocks online, I set
out to build a shopping list to understand how everything is bought and
sold within this black market. It all started with checking sources through a
few Web searches then making trips into Web forums I dared visit only
with a virtual machine and Google Translator’s help. All I had to do was
paste in “botnet” in Cyrillic, and I was on my way down the rabbit hole.
To assemble your list for some of the simplest get-rich-quick schemes, all
you need is about $600, a little spare time, and no compunctions about
breaking laws to make a profit. I didn’t deploy an Ars-enal of botnet
destruction in the end, but I absolutely could have. That may be the
scariest lesson here.
It looks like you’re trying to build a botnet…
There are no personal shoppers to help walk you through the underground
marketplaces to identify what fits a particular criminal scheme—though
there may be plenty of people willing to give you paid advice on how to
get started. With absolutely no budget for bitcoins, I got my start with
some help from Max Goncharov, a security researcher for Trend Micro
who specializes in following the Russian underground marketplaces for
online fraud services. Goncharov came to Washington, DC in late March
for a Trend Micro press briefing, and he laid out some of the basic things
that go into a beginner fraudster’s software and services shopping cart:
botnets, malware-spreading tools, and hacking for hire. (Goncharov
detailed some of these services in a paper published late last year and
presented during this press road show.) Goncharov’s suggested setup came
with a $595 price tag for the first month of operations and a monthly cost
of $225 to sustain the operation.
Many of the sites run under the thin veneer of “security” discussion
boards. But they’re often paid for by advertisements for the tools sought
by a certain class of cyber-criminal: botnet-herders and the service
provider ecosystem that has sprung up around them. These are largely the
small and medium businesses of cybercrime, following a well-worn
approach to making money. If you cast a big enough net, you’re bound to
catch some fish.
The botnet herders’ standard business plan is to “use exploit kits, and then
run a phishing campaign or some sort of campaign against massive
numbers of people with hopes that someone is going to click on a link and
get the exploit to drop a botnet or banking trojan onto their machine,” said
Nicholas J. Percoco, senior vice-president of Trustwave and head of the
company’s SpiderLabs penetration testing and security research team.
“Once they’ve done that, it goes down the path of them monitoring them
when they do banking transactions, or the botnet may be involved in spam
or distributed denial of service attacks. Or maybe it’s a sort of Swiss Army
knife botnet that can do many different things depending on what that
botnet herder decides, or what he makes it available to do for people who
want to utilize his or her botnet.”
No matter what the racket, Percoco told Ars, the equation for botnet
herders is the same. “From a criminal’s perspective, they’re looking at
massive numbers of attacks to achieve their financial goals.”
Botnet software itself is an important part of the whole equation, but it’s
only a fraction of first-month startup costs—and one that’s fungible if
you’re willing to invest some of your own sweat equity in the setup (or
dispense with the “legitimate” route and use a cracked version without
software support). Just like any Internet business, launching a financial
fraud botnet—or any kind of long-running botnet endeavor—requires a
sustainable business plan. You need to know your target market, ensure
distribution, keep your installed base a step ahead of the competition, and
keep your business processes secure.
Before you start building a bot army of incredible magnitude, you need—
just as with any other hacking endeavor of questionable legality—to hide
yourself from prying eyes. That means using some sort of tool to avoid
monitoring by your ISP, law enforcement, and other cybercriminals.
Generally speaking, the best way is a virtual private network.
As confessed LulzSec member Cody Kretsinger found out, not all VPN
providers are created equal. He used a service called HideMyAss.com, a
VPN and proxy service run by UK-based Privax Ltd. Unfortunately, he
didn’t read the company’s privacy and legal policies, and they gave up his
logs when law enforcement came knocking. “Bulletproof” VPN services
are ones that claim to be shielded from law enforcement requests because
of their location or logging practices. Many of these services have
disclaimers about “abuse” of the services, but the fact is that they take a
number of anonymous forms of payment (CryptoVPN, for example,
accepts Liberty Reserve, Bitcoin, and a number of other similar
anonymous payment services). At worst, these services may just cancel
your account if it attracts too much trouble.
Once you’ve got your network secured, you need some place to host your
botnet’s command and control network and all of the other assorted
badness needed to launch a massive assault on the unsuspecting world. For
those without the skills, time, or desire to simply go hack someone else’s
server every couple of weeks, that means buying a dedicated or virtual
dedicated server from someone who doesn’t care what you’re doing—lest
your botnet’s nerve center be wiped during a security sweep or seized by
law enforcement.
All of this, the company says with a wink, takes a lot of time and human
resources, and as a small company there may be some delays before it gets
around to it. In other words, “don’t worry—we’re inefficient.” Hostim
VSE’s dedicated servers start at $39/month with additional charges for
more bandwidth. The company also, ironically, provides DDoS protection
and other support services. But prices for its services will rise dramatically
if you’re attracting too much attention or using too much bandwidth.
For really hard-core criminal undertakings, there are the more specialized
underground “bulletproof” hosting services that are run specifically for
malware owners. These offer hosting at a significant markup in exchange
for looking the other way. These operations generally don’t maintain
webpages. They advertise strictly in underground forums and do business
over ICQ, Jabber, and other instant messaging.
Mihai Ionut Paunescu, the 28-year old Romanian behind underground host
powerhost.ro, was caught in December by Romanian authorities. His
servers were home to the Gozi financial malware command and control
network. Paunescu kept tabs on exactly what sort of business his users
were up to and charged accordingly. In some cases, the rates reached
thousands per month, averaging better than a 100-percent margin on the
servers he managed.
You’ll also want to register those domains with a registrar that’s not going
to roll on you and shut you down on the first complaint of abuse. You need
someone who will shield your identity from the prying eyes of security
firms and law enforcement. In other words, you want a bulletproof
registrar. Preferably, it’s one that accepts payments via Western Union or
some other anonymous service.
Of course, it’s never a bad idea to have some additional protection to make
sure that you cover your trail completely by using a “fast flux” scheme.
This hides your servers’ true location by assigning the DNS addresses to a
rapidly changing set of proxies. Fast flux providers will take your domains
—or even register them for you—and then assign host names to a
collection of their own bots. These in turn pass traffic between your bots
and your server. By using a short “time to live” for the host “A” name
records in the DNS server, fast flux systems create hundreds of potential
communication paths for the bots.
Do-it-yourselfers who don’t care about things like patches and full support
can find “cracked” versions of some of these toolkits (or ways to disable
their licensing code) for free. There’s also an option to pick up older but
still supported versions on the aftermarket. Zeus’ source code was released
to the world last year, sort of turning it into “open source,” so you can now
purchase supported versions of it for $125, plus $15 per month for updates
to the code and $25 for monthly 24/7 new customer support. That could
include everything from helping a noob fix a misconfigured server to
doing a whole walkthrough of configuring the PHP scripts and MySQL
backends for the system over Skype. You could always just use one of
those YouTube guides, though, and save some money.
There are financial botnet injectors that insert Web code into banking sites.
These injectors try to grab your personal information to hijack your
account, make wire transfer withdrawals, or even change the values
presented in your online statement to conceal all of it.
Beginners can buy injector packs for a set of banks through marketplaces
and pay for direct support to help install, tune and customize them. In
some cases, these require some server setup as well to properly harvest the
data collected. It sounds complicated, but there are people happy to help
you figure it all out for a small fee.
Budget Botnet Shopper’s Price: $80, plus $8/month for support.
An exploit tool or service
The modus operandi of most bot herders is to use Web links as the delivery
method for their malware, sending out streams of spam to potential victims
in the hope that someone will fall for their social engineering. One click
leads to a webpage set up to drop a package of nastiness on them. There
are ready-made exploit packs, loaded with code written specifically for the
purpose of planting malware on victim’s PCs that can be purchased and
installed on a Web server or “rented” as a service.
Botnet builders can rent capacity on these services or outright buy them. A
Phoenix exploit kit (like the one used to seed the recent Bamital botnet),
can be purchased for $120, plus another $38 per month for patches and
technical support, according to Goncharov. BlackHole, another market
leader in exploits, offers its latest and greatest as a leased service for $50 a
day, with extra fees for traffic overages. BlackHole also comes with an
Oracle-like annual license for those who want to deploy on their own
server. That costs $1,500 per year, with various add-on functionality fees.
Budget Botnet Shopper’s Price: $120 for the kit, plus $38/month for
support.
Crypters and dropper builders
The problem with just pushing a Zeus bot in its raw form out to targets
through an exploit is that the Zeus bot is bound to be detected by antivirus
software because of its signature. To prevent that, botnet-herders turn to
“dropper” malware that disguises the bot trojan, delivering it in encrypted
form to disguise the file signature of the trojan and its associated files.
There are even crypter services now available on the Web, delivered as a
service. One such service offers dropper-building at the rate of $7 per
“sample.”
Another key to not getting caught is “antisandbox” code that detects if the
malware has been dropped onto a sandboxed system or virtual machine—
such as those used by digital forensics experts and security analysts. If the
code detects that it’s been deposited inside such a system, it can prevent
the botnet trojan from being deployed and giving up the nature of the code
it carries. Antisandbox code is a feature of some crypters.
But wait! You still have to deliver the exploit link to drop your botnet
package on your unsuspecting (or possibly suspecting) victims. How do
you get them to click on that link? The traditional route is by blasting
semi-convincing spam messages and hoping people are dumb enough to
click on a link in them to see a video, download a document, or reconfirm
their PayPal information.
That typically means buying a spam blast, often from another botnet
operator. Some spam-masters have moved their focus to social networks
and charge not per message but per hit. You can spam out to social
networks and over SMS with clickjack links using “borrowed” credentials
or leave it to the professionals to do it for you for around a buck per
thousand targets.
But that’s the old-fashioned way. The new-fashioned way is to use “spear-
phish” attacks that use social information about the target in some way that
convinces them to click the link, either through a social network message
in a compromised account or an e-mail that appears to be from a friend. If
you want to make it even more convincing, you can always pay someone
to hack a victim’s e-mail address to get access their account and contacts.
Then it’s a simple as posing as the victim to fool all their friends.
For beginners, spam remains the best bet. It can be used to hit a variety of
potential targets and it’s relatively cheap: cheap spamming services can
run as little as $10 per 1 million e-mail addresses, with better services
based on stolen customer databases running five to ten times as much.
Using our budget shopper prices, that adds up to about $576 for the first
month of operation.
On the upside, it’s an easy game to buy into—unlike the bigger, more
enterprise-scale cyber-crime rings behind big corporate data breaches.
While the whales of the cybercrime game may share some of the basic
technology approaches with their smaller cousins, they have more in
common with the intellectual property stealing “Advanced Persistent
Threat” (APT) hackers alleged to be associated with the Chinese military
(though they may surpass them in skill). Trend Micro Chief Technology
Officer Raimund Genes said during the DC briefing that he thought the
recent alleged Chinese APT attacks had been uncovered largely because
they lacked the finesse of Eastern European cybercrime rings.
As a result, the data breaches caused by these targeted hacks can go on for
months, even years before being detected. A study released by Percoco’s
team at Trustwave in February found that the average targeted attack went
more than 210 days before it was detected. And this detection was usually
because of a customer complaint or notification by law enforcement or a
payment processor, not because antivirus software detected the hack. At
some companies, the hacks lasted more than three years without being
detected, all while millions of credit card transactions and other data were
being pumped back to the hackers.
Botnet operators generally go big or go home in their attacks. But the tools
they use can just as easily be applied to the long game if they’re used in a
targeted fashion and they apply some of the lessons learned by the big-
time hacking organizations. “Swiss Army knife” botnets and remote
administration tools can be used as part of a poor man’s APT by those who
are willing to take the time to do the research and social engineering to get
their malware in the right place. And just because Zeus and other botnets
are a known threat doesn’t mean they can’t be used in stealth. According to
the siteZeusTracker, the average detection rate for Zeus binaries by
antivirus software is only 38 percent. And that’s for known Zeus botnets.