As someone who's worked in the financialindustry for years, I'm fascinated by methods used by phishers to encourage people topart with their money. Most of us can easily recognize and avoid the more obviousand clumsy phishing attacks (419 scams, emails in broken English claiming to befrom "PayPaI," etc.), but how equipped are we do deal with the threat of rogueIVR systems?
IVR (Interactive Voice Response) is a technology that letshumans communicate with computers over a phone line, using voice recognitionor touch-tones from a phone's keypad. We've all had experience with IVR systems:the automated attendants and phone menus we love to hate are implementations ofIVR technology.
Your financial institution's promise to never ask for yourpersonal information in an email doesn't really stop determined phishers; it just forcesthem to be more creative. That's where rogue IVR systems come in. Phishers cantrick their victims into entering their personal information by emulating thelegitimate IVRs of real financial institutions.
Building your own IVR is actually quite simple. TakeAsterisk, for instance. Get the hardware, get the software, install andconfigure Asterisk on your Linux machine, and you've got yourself an IVR, readyto go. Get a cheap 800 number and you're two-thirds of the way to ruiningsomeone's day.
The general attack model goes something like this. Usingtheir own IVR systems, attackers send victims emails or SMS text messages nearlyidentical to real financial institution alerts regarding fraud detection orsimilar. The fake message is careful to substitute the phone number of thefinancial institution with the attacker's number, however. The victim calls theprovided number and enters their personal information when prompted by the fakeIVR.
The lazy phisher won't care how the real financialinstitution's IVR works; he'll just prompt the victim to enter theirinformation but reject their credentials every time. This may net him severalpasswords and PINs, but it will raise red flags with victims when they realizetheir credentials aren't working when they should.
A somewhat more ambitious phisher will call the financialinstitution's real number and record the prompts and responses given by theIVR's automated system. He then uses this recorded audio to add an element ofprofessional believability to the attack.
Now let's take a look at the most transparent (and most impressive)implementation of a rogue IVR. At DEF CON 20, this attack was demonstrated to aroom full of hackers at the Social Engineering CTF. The first part of theattack is the same as above: get the victim to call your IVR by dialing your number insteadof their financial institution's number. But when the call comes in, make your IVRdivert the call to – wait for it – their actual financial institution! Thevictim will hear legitimate prompts coming from the legitimate IVR and enterhis personal information – while your IVR captures that information as it plays man-in-the-middle.Unless the victim sees his financial institution's phone number later andrealizes he dialed something different, he will have no idea he's been had.
The moral of this story? The only way to consistently defeatphishers and their rogue IVRs is to neverdial a phone number provided in an email or SMS, even if it looks legit.Instead, save your financial institution's phone number to your contacts listand dial it from there.