We dropped in on [Charlie Miller]‘s fuzzing seminar at the end of the day yesterday. Fuzzing become a fairly popular topic in the last year and essentially involves giving a program garbage input, hoping that it will break. If it can’t handle the fake data and fails in a non-graceful fashion, you could have found a potentially exploitable bug. Fuzzing is a fairly simple idea, but as Charlie points out, without some thinking while you’re doing it it’s unlikely to be very productive.
Whenever someone manages to expose vulnerabilities in everyday devices, we love to root for them. [Adrian] over at Irongeek has been inspired to exploit barcodes as a means to attack a POS database. Based on an idea from a Pauldotcom episode, he set out to make a rapid attack device, using an LED to spoof the signals that would be received by scanning a barcode. By exposing the POS to a set of generic database attacks, including XSS, SQL Injection, and other errors easily solved by input sanitation, he has created the first version of an automated system penetration device. In this case the hardware is simple, but the concept is impressive.
With the hardware explained and the source code provided, as well as a basic un-sanitized input cheat sheet, the would-be barcode hackers have a great place to start if they feel compelled to provide a revision two.
[Thanks Robert W.]
Here’s a quick prototype from [Travis Goodspeed]. It’s a smart card built around an MSP430 microcontroller. We’ve used the MSP430 in the past because of its low power demands. He says this business card currently supports 1.8V to 3.3V, but a future design will have 5V as well. Technologies like Java Card exist for running applets on smart cards, but a familiar microcontroller like the MSP430 could certainly make development much faster. Knowing [Travis], there’s a reader somewhere about to go through some serious fuzzing.
[Jeremiah Grossman] and [Eric Lawrence] will be presenting on clickjacking and browser security in an online seminar tomorrow. Clickjacking allows an attacker to transparently place links exactly where a user would be clicking, essentially forcing the user to perform actions without their knowledge. This method of attack has been known for a few years, but researchers have focused their attention on it lately because they feel the threat has been underestimated. Recently, Adobe patched a vulnerability specifically because of this issue. Tune in tomorrow for more info on the attack.
T-Mobile’s G1 was released last week and there has been at least one Android vulnerability announced already. The New York Times reported on research done by [Charlie Miller], who also helped find one of the first iPhone bugs, so we think the report is fairly credible. Last year, we saw him deliver a seminar on real world fuzzing at ToorCon 9. It covered exactly how they found the iPhone bug.
If you just want to use a G1 without service, you can activate it with any T-Mobile SIM card.
The iphone-dev team published a video today showing access to the iPhone’s baseband processor. They connect to the device over ssh and then use minicom to issue AT commands. They’re writing custom AT commands for full control.