Botnet Billionaires | TechSNAP 170

chris

8 years ago

Want to make billions in days? Quit your job and become a botnet master. We’ll share the story about a Brazilian botnet that you’ve just got to hear.

Plus a major flaw in Android, encryption done right, your questions, our answers & much much more!

Thanks to:

Direct Download:

RSS Feeds:

— Show Notes: —

Botnet stealing from Brazilian banks rampent, maybe into the billions of dollars

In Brazil, the most common form of payment, for everything from taxes, utility bills or large purchases and almost all business-to-business payments is “Boleto Bancario” (or just boleto for short)
It is basically an bank transfer, somewhere between a cheque and a wire transfer
Most Brazilians do not have credit cards, and credit card processing is expensive (usually 3-5% or more) and the merchant usually has to wait 30 days to receive the funds
A boleto usually only takes 24 to 48 hours and has a low fixed fee (approximately $1)
unlike credit card payments, which can be disputed and reversed, boleto cannot be reversed. Refunds are handled by bank transfer
The information is filled out on a form, and then the recipient enters the details online to receive the payment
Brian Krebs was shown a botnet that was lying in wait on infected computers, and as the user entered the details of a boleto, it would quickly change the recipient as the transfer was submitted, allowing the botnet controllers to receive the money, instead of the intended recipient
“Thieves had hijacked some 383 boleto transactions between February 2014 and the end of June, but had stolen the equivalent of nearly USD $250,000 during that time”
Researchers at RSA Security (part of EMC) found an even larger botnet
“RSA says the fraud ring it is tracking — known as the “Bolware” operation — affects more than 30 different banks in Brazil, and may be responsible for up to $3.75 billion USD in losses. RSA arrived at this estimate based on the discovery of a similar botnet control panel that tracked nearly a half-million fraudulent transactions.”
“Most Brazilian banks require online banking customers to install a security plug-in that hooks into the user’s browser. The plug-ins are designed to help block malware attacks. But according to RSA, the Bolware gang’s malware successfully disables those security plug-ins, leaving customers with a false sense of security when banking online.”
“RSA notes that the miscreants responsible for the Bolware operation appear to have used just over 8,000 separate accounts to receive the stolen funds.”
The botnet Krebs discovered was much less sophisticated, using only 3 destination bank accounts
RSA PDF

Dealing with encrypted streams

Adam Langley (of Google Security, and one of the authors behind BoringSSL) posts on his blog about how many file encryption systems, including gnupg, get it “wrong”
Specifically, when encrypting large messages they often use a single MAC (Message Authentication Code) at the end of the message
A MAC is used to ensure that the ciphertext has not been modified or corrupted before attempting to decrypt it
The problem is, if you do something like this: gpg -d your_archive.tgz.gpg | tar -xz
It will decrypt the contents of the gpg encrypted file and spit them out to the pipe, and not until it reaches the MAC at the end of the message, will it realize that the file was corrupted, and should not have been used. At this point it is too late, tar has already processed the invalid stream
An attacker may be able to use this to cause tar to overwrite a file the user did not intend, or otherwise create corrupted files or exploit a vulnerability in tar
The correct way to handle this situation is to not return the data until it has been authenticated, however this may require an impossibly large buffer
The author discusses the reasonably low overhead (0.1%) of breaking the message into 16 KiB chunks, each with a 16 byte MAC. This would allow gpg to authenticate each small chunk before writing it to the pipe.
However, with that approach “Although safer in general, when chunking one has to worry that an attacker hasn’t reordered chunks, hasn’t dropped chunks from the start and hasn’t dropped chunks from the end”
Ted Unangst (of OpenBSD/LibreSSL) posts his thoughts
Ted clarifies that OpenBSD’s ‘signify’ system in newer OpenBSD installers download the archive, verify the downloaded temporary archive before passing it to tar to be extracted, as opposed the the old design before signify, where the file was piped to tar directly from the ftp client, not requiring the temporary storage space
Ted also mentions his ‘reop’ (Reasonable Expectation Of Privacy) tool, a light weight (incompatible) replacement for GnuPG, “However, the entire message must decrypt and authenticate successfully before any output is produced, so it’s actually safer than a small packet streaming program which may produce partial output. (reop cheats a bit by imposing a message size limit; it simply can’t encrypt large files, for large values of large.)”

Android keystore stack overflow flaw could allow key-theft

The vulnerability could allow attackers to steal cryptographic keys from the device, including those for some banking services, virtual private networks, and PINs or patterns used to unlock vulnerable devices
The flaw is fixed in Android 4.4
Originally incorrectly reported as affecting 86% of devices, it only affects ~ 10.3% as it only affects Android 4.3
The vulnerability requires a malicious app be installed on the targeted handset, but we have seen legitimate apps be bought or hijacked before, and it is often fairly easy to trick people into installing apps
“Generally speaking this is how apps are going to store their authentication credentials, so if you can compromise the KeyStore, you can log in as the phone’s user to any service where they’ve got a corresponding app, or, at least, an app that remembers who you are and lets you log back in without typing a password. This means that most banking apps, which force you to type your password every time, are probably safe against this particular attack.”
Researcher Post

Feedback:

Round Up:

Banking Malware Intercepts and Logs Network Traffic – Sent in by Mark
Unix wildcards gone wild
How to watch hacking, and cyberwarfare between the USA and China, in real time
Norse – IPViking Live
Bug in Amazon Fire TV causes Screensaver to use 80GB per day of data, blowing users’ transfer caps
Goldman Sachs demands Google “Unsend an emails” that is accidently sent to @gmail.com instead of @gs.com. Google said “not without a court order”, Goldman Sachs filed with the New York Supreme Court, requesting “emergency relief” to avoid a privacy violation and “avoid the risk of unnecessary reputational damage to Goldman Sachs.”
Microsoft kills “Patch Tuesday” emails, blames new Canadian anti-spam law
- Microsoft provided feedback and was involved in every step of drafting the Canadian anti-spam law. It can as quite a shock to all involved when Microsoft announced they would be killing off their email services. After the uproar, Microsoft quickly reversed course “On June 27, 2014, Microsoft notified customers that we were suspending Microsoft Security Notifications due to changing governmental policies concerning the issuance of automated electronic messaging. We have reviewed our processes and will resume these security notifications with our monthly Advanced Notification Service (ANS) on July 3, 2014.”
Fallout from the poor communications surrounding the announcement of a 2012 study by facebook which saw it modify the tone of stories displayed to users to track how it changed the emotions of the users. Surprise: Showing users sad stories made them sad, and vice versa
NZ Court rules Dotcom encryption keys cannot be given to FBI
Medal for writing Perl for the US Army
Malware targetting ICS may have been able to sabotage electrical grid
- “Its most vicious attack campaign saw it imperil several industrial control system (ICS) equipment providers, infecting their software with a remote access-type Trojan. This resulted in companies installing the malware while downloading software updates for computers operating ICS equipment. These attacks not only gave the hackers a foothold in the targeted companies’ networks, but also provided them the means to mount sabotage operations against infected ICS computers”