The name “iBOT” is already taken. It may soon be taken in vain if the iPhone becomes a popular target for malware.
For years Ive been continuously incredulous at the lack of interest by the black hat community in the Mac platform. There have been large numbers of serious vulnerabilities, many of which sat unpatched for months after being disclosed. Consider the Month of Apple Bugs, one of the best in the “month of xxx bugs” genre.
Click here to read more about the Month of Apple bugs.
Now the opportunity for attack on the iPhone is growing. Thanks to HD Moore, the indispensable man both for good guys and bad guys in the security business, we have a version of Metasploit for the iPhone. We also know, thanks to him, that Apple iPhone processes all run as root.
Im not sure whether to be surprised at the “running as root” revelation. There could be good reasons why Apple did it, probably performance-related. But the fact is that it is a problem for the same reason that running Windows as Administrator is a problem: Once malicious code gets to run in the context of any of those programs, your iPhone is completely 0wned.
How would such code get to run? There is no formal user-accessible interface for running native programs on the iPhone. Apples idea is that third-party apps should run in the browser. But hackers have created many native apps for it and package managers (such as Installer.app) for installing them.
How did they get these tools on in the first place? Initially, probably through telnet and other complicated mechanisms. There are also hacks using iTunes and a system called “jailbreaking,” but there is no easy way to do it accessible to normal users.
So how do hackers do their dirty work if typical users dont normally install non-Apple apps? The answer is vulnerabilities. All you need is a malicious Web page that you advertise through spamming, probably to Mac-heavy domains like mac.com, apple.com, and certain universities. This basic architecture is the mainstream in Windows malware these days.
The iPhone 1.0.1 update fixed a vulnerability that sounds like it fit the bill: CVE-2007-3944, in which, the update says, “viewing a maliciously crafted Web page may lead to arbitrary code execution”.
Whether this vulnerability really allows for complete compromise isnt really the issue. After all, its patched. But obviously there are others. The iPhone software is complicated enough that it must have many vulnerabilities; Safari has a long history of them. Many will go unpatched for long periods, and those that get patched will not get to all phones quickly because you have to go to iTunes to see if there are patches available.
That leaves two questions: Would attackers really go after iPhones, and what would they do with them? As I said above, Mac users have always been safe from attack, not because of a safer platform but because nobody was bothering to attack it. Perhaps the iPhone is a more attractive target.
There has also been a long history of mostly theoretical mobile phone attacks, mostly on the Symbian platform. For years security companies and analysts have said that a flood of such attacks was imminent. Heres just one that has been found in the wild on other phones: a dialer scam. Use the phones to initiate calls to Third World countries with large—but not too large—charges on the account. There are a lot of iPhones out there; you could make a lot of money this way.
AT&T is also at risk. Since all (unlocked) U.S. iPhones are on its network, the potential exists for the iPhone to be used specifically for attacks on that network. This is the unfortunate flip side of its monopoly on support for the device.
Security Center Editor Larry Seltzer has worked in and written about the computer industry since 1983.
More from Larry Seltzer
Check out eWEEK.coms for the latest security news, reviews and analysis. And for insights on security coverage around the Web, take a look at eWEEK.com Security Center Editor Larry Seltzers blog Cheap Hack