Shaun Stanislaus’s Tech blog

Just another WordPress.com weblog

Twitter spoofing: The next logical exploit

First it was spoofing e-mail, then IM, and now spoofing Twitter is the new means of exploit. How attractive really is the ROI for attackers?

I just completed an article titled “URL shortening: Yet another security risk“, in which I discussed URL shortening and how phishers/attackers subverted it to drive unsuspecting users to malicious Web sites.

After reading the many comments, I was happy to note that in general users are getting savvier about misdirection exploits.

This appears to apply to Twitter as well, even though messages or tweets, with shortened links make it more vulnerable.

Fortunately, Twitter has an additional advantage in that we the users get to pick who can send us tweets. This capability significantly reduces the risk simply because you know who’s sending you the message.

Well, maybe not
I’ve just finished reading an article by Washington Post’s Brian Krebs titled “Twitter security hole left accounts open to hijack“. It seems that it’s not that difficult to spoof Twitter messages.

Krebs quoted Lance James a security researcher and author of “Phishing exposed“:

“Anyone could authenticate and hijack a Twitter account by using SMS spoofing services, such as my-cool-sms.com, or phonytext.com. These Web sites allow users to mask what phone number they are texting from by letting the user input whatever phone number they want to appear in the from field.”

Oh great, this totally negates the one advantage that Twitter had over IM and e-mail. It’s not hard to see that phishers/attackers would want to leverage SMS spoofing along with URL shortening to redirect victims to malicious Web sites.

Help from the cellular network operators
One good thing that Krebs alluded to was the fact that SMS spoofing may only work if the attacker is located outside of the United States:

“Twitter co-founder Biz Stone wrote in an e-mail.[Mobile] carriers in the U.S. have their own systems for blocking SMS spoofing. Indeed, most U.S.-based mobile carriers have put in place measures to block SMS spoofing on their networks. But this is generally not the case for international mobile networks.”

It appears that United States is one of the few countries forcing cellular carriers to clamp down on SMS spoofing. That’s great, but spoofing Twitter messages is still possible just about everywhere else. I’ll give you two guesses where most phishing and malware exploits originate, and the first one doesn’t count.

Proof of concept
H Security (a German security company) verified that SMS spoofing works in an article titled “Twitter spoofing fix fails in UK and Germany“. The article provides the following details of the process:

“In the UK, we had a mobile phone associated with a Twitter account. By taking only the number of the mobile phone and setting it as the sender field on PhonyText then sending an SMS to +447624801423, the UK number for sending SMS tweets, we were able to see our message appear in the tweets on the honline page.”

The article goes on to explain what this potentially means:

We then promptly removed the association between the phone and the Twitter account. An attacker could have created a message directing followers to malware sites, to other risky locations on the web, or posted tweets designed to ruin the reputation of the account owner.”

What this means
First, the ability to spoof a Twitter message enhances all the normal misdirection schemes that are already in play. The fact that shortened URLs are common place in Twitter messages makes it even easier to pull the scheme off.

The damages from the SMS spoofing and URL shortening exploit can be as simple as malware being loaded on victims’ computers to as complex as stealing sensitive financial information from the victims. Also a cruel joke could be played on Twitter accounts that don’t have unlimited texting. It would be easy to run up some monster phone bills as noted in the Twitter support section:

“Twitter charges you nothing, but how much it costs to use Twitter with text messaging depends on your text messaging plan. Standard text messaging rates (such as international text messaging fees) do apply. Consult your service provider to ensure that your text plan covers your Twitter usage.If you’re using our international number, give your provider the Twitter phone number you’ll be using to see if you’ll incur extra charges. If you’re using Twitter from outside of the US, please consult your carrier, as every provider has a different policy.”

Final thoughts
Following spoofing’s logical progression was easy for the phishers and malware creators of the world. Yet, from the comments I’ve read, it seems like it’s getting harder for them to find chinks in the armor. That’s good and should be heartening to all of the people who are trying to keep the Internet the amazing place it is.

Still, there needs to be awareness and vigilance as long as the possibility of a ROI is perceived by the dark side.

Advertisements

April 15, 2009 Posted by | hacker, IT News, Security | , , , , , , , , , , , , , | Leave a comment

Conficker’s autorun and social engineering guide

We wrote several diaries about Conficker (or Downadup, depending on the AV tool you are using). F-Secure posted some interesting information about the number of infections which is almost certainly in millions (and who knows how many machines will stay infected as the owners will not even notice anything).

One of the reasons for infecting so many machines is that Conficker uses multiple infection vectors:

  1. It exploits the MS08-067 vulnerability,
  2. It brute forces Administrator passwords on local networks and spreads through ADMIN$ shares and finally
  3. It infects removable devices and network shares by creating a special autorun.inf file and dropping its own DLL on the device.

F-Secure also blogged about the autorun.inf file where they noticed that it contained a lot of garbage (about 60 kb of random binary data). This fooled some AV programs so they didn’t scan the device properly (otherwise, they would have picked up the referenced DLL also stored on the device).

After removing garbage, one can see a nice autorun.inf file containing all important keywords. This grabbed my attention:

[Autorun]

Action=Open folder to view files
Icon=%systemroot%\system32\shell32.dll,4
Shellexecute=.\RECYCLER\S-5-3-42-2819952290-8240758988-879315005-3665\jwgkvsq.vmx,ahaezedrn

So, as you can see, the first part, “Install or run program” is there because Vista detected an autorun.inf file containing the shellexecute keyword. However, the text comes from the Action keyword and the icon is extracted from shell32.dll (the 4th icon in the file) – and it’s the standard folder icon! This can easily fool a user in clicking this one and thinking it will open the USB stick in Windows Explorer instead of the second (the real one). The first option will run Conficker, of course. Very smart. For administrators among you, I would suggest that you disable AutoPlay in your environments, unless it’s really necessary. Depending on the environment you might even completely disable USB, if you don’t need it. The following article explain nicely how the AutoPlay feature works and how to disable it (http://technet.microsoft.com/en-us/magazine/2008.01.securitywatch.aspx). Or check this article on the Autorun registry key (http://support.microsoft.com/kb/953252). UPDATE – fixed a typo in the vulnerability, it is MS08-067 (not MS08-068) – Nick Brown sent a URL to his blog where he described another method for disabling Autorun by modifying the IniFileMapping registry key, see more at http://nick.brown.free.fr/blog/2007/10/memory-stick-worms.html

April 2, 2009 Posted by | social engineering | , , , , , , , , , , , | 1 Comment