Satellite television is prevalent in Europe and Northern Africa. This is delivered through a Set Top Box (STB) which uses a card reader to decode the scrambled satellite signals. You need to buy a card if you want to watch. But you know how people like to get something for nothing. This is being exploited by hackers and the result is millions of these Set Top Boxes just waiting to form into botnets.
This was the topic of [Sofiane Talmat’s] talk at DEF CON 23. He also gave this talk earlier in the week at BlackHat and has published his slides (PDF).
The Hardware in Satellite receivers is running Linux. They use a card reader to pull in a Code Word (CW) which decodes the signal coming in through the satellite radio.
An entire black market has grown up around these Code Words. Instead of purchasing a valid card, people are installing plugins from the Internet which cause the system to phone into a server which will supply valid Code Words. This is known as “card sharing”.
On the user side of things this just works; the user watches TV for free. It might cause more crashes than normal, but the stock software is buggy anyway so this isn’t a major regression. The problem is that now these people have exposed a network-connected Linux box to the Internet and installed non-verified code from unreputable sources to run on the thing.
[Sofiane] demonstrated how little you need to know about this system to create a botnet:
- Build a plugin in C/C++
- Host a card-sharing server
- Botnet victims come to you (profit)
It is literally that easy. The toolchain to compile the STLinux binaries (gcc) is available in the Linux repos. The STB will look for a “bin” directory on a USB thumb drive at boot time, the binary in that folder will be automatically installed. Since the user is getting free TV they voluntarily install this malware.
Click through for more on the STB Hacks.
Continue reading “Millions of Satellite Receivers are Low-Hanging Fruit for Botnets”
Looks like the FBI is starting to get pretty serious about fighting malware. Traditionally they have attacked the servers that activate and control botnets made up of infected computers. This time they’re going much further by taking control of and issuing commands to the botnets. In this instance it’s a nasty little bug called Coreflood, and they’ve been given permission to take the yet-unheard-of step by a federal judge.
An outside company called Internet Systems Consortium has been tapped to do the actual work. It will call upon the malware on infected computers and issue a command to shut it down. That falls short of fixing the problem as Coreflood will try to phone home again upon reboot. This gets back to the initial problem; we won’t ever be able to stop malware attacks as long as there are users who do not have the knowhow (or simply don’t care) to protect and disinfect their own computer systems.
How long do you think it will be before some black hat comes up with a countermeasure against this type of enforcement?
The folks over at Arbor Networks were browsing Twitter and discovered something very strange: a Twitter account seemingly posting gibberish. At least, that’s how it appeared at first. Upon closer investigation, they discovered that the profile was posting base64 encoded links to PKZIP archives. When they extracted the contents and unpacked the contained DLL and EXE files, they discovered that the account was posing links to malware that would post user information back to certain URLs. The article was also updated to show that the scheme wasn’t limited to Twitter, but also affected users on Jaiku and Tumblr. It’s a bit scary to see that all malware isn’t as blatantly obvious as we usually would think it to be.
PandaLabs has identified a botnet running a malware campaign impersonating president-elect Obama’s website. The front page of the site features a sensational story titled “Barack Obama has refused to be a president”. Clicking the link will download the malware and make the target’s machine part of the botnet. They’re using fast-flux to assign the malicious domains to the massive number of compromised nodes that are hosting the actual site. The team has contacted the domain name registrar in China to get the domains removed. Using a sensational headline is not new to malware; it’s how the Storm Worm got its name.
Zero Day has an interview with German researchers who have found a way to take down the Storm Worm botnet. Their program, Stormfucker, takes advantage of flaws in Storm’s command network: Nodes that are NAT‘d only use a four-byte XOR challenge. Nodes that aren’t NAT’d are only using a trivial 64bit RSA signature. Their solution can clean infected machines and also distribute to other nodes. Unfortunately, installing software without the user’s consent is the exact same behavior as malware. Don’t expect to see this in any sort of widespread use. The researchers did point out that some ISPs have moved to shutting off service for infected customers until their machines are cleaned.
The Washington Post is reporting that the shutdown of one hosting company has caused the total volume of spam to drop by 2/3rds. The company in question is McColo Corp. Both Hurricane Electric and Global Crossing pulled the plug today after a damning report revealed a number of illegal activities happening on McColo’s servers. McColo already had a reputation with the security community. When contacted about abuse, the company would often shift servers to new IP ranges instead of shutting them down. Although not the main source of spam, the company was host to many botnet control servers and phishing sites.
P2P networks have long been a legal gray area, used for various spam schemes, illegal filesharing, and lots and lots of adware. Last year, though, the first botnet created by a worm distributed via P2P software surfaced, the work of 19-year-old [Jason Michael Milmont] of Cheyenne, Wyoming, who distributed his Nugache Worm by offering free downloads of the P2P app Limewire with the worm embedded. He later began distributing it using bogus MySpace and Photobucket links shared via chats on AOL Instant Messenger. The strategy proved effective, as the botnet peaked with around 15,000 bots. [Milmont] has plead guilty to the charges against him. Per his plea agreement, he will pay $73,000 in restitution and may serve up to five years in prison.