Despite a few security hiccups, Apple's OS X and iOS platforms have been generally very secure. While all users are potentially vulnerable to phishing attacks and identity theft, Apple's platforms have been solid. But that seems to be changing with the identification of a flaw in Apple's Thunderbolt ports that can used to write custom code into the Mac's boot ROM.
Dubbed Thunderstrike by engineer Trammell Hudson, the attack works by installing a rootkit via a Thunderbolt port by using an infected host. Once the rootkit is installed, it can infect other devices as it can then use devices using the Mac's internal Thunderbolt interface. This allows t to propagate between devices.
The threat works because a programmer can use a Thunderbolt Option ROM to circumvent the cryptographic signature checks in Apple's EFI firmware update routines. An attacker can write untrusted code to the SPI flash ROM on the motherboard. As there aren't any cryptographic checks of firmware validity at boot time, the malicious code controls the system from the very first instruction and hide itself from detection.
According to Hudson, once installed, Thunderstrike will persist even after the OS is reinstalled. Since the boot ROM is independent of the operating system, reinstallation of OS X will not remove it. As it doesn't depend on anything stored on the disk, replacing the hard drive has no effect. The only way to remove it is with another Thunderbolt device that restores the ROM to its correct configuration.
At this stage, injection of the Thunderstrike requires either a user to connect an infected device or a motivated actor to physically access the device. Hudson's proof of concept bootkit, which was demonstrated at the recent Chaos Computer Club conference, also replaces Apple's public RSA key in the ROM and prevents software attempts to replace it that are not signed by the attacker's private key.
Hudson says "that this attack uses can be closed with a few byte patch to the firmware, the larger issue of Apple's EFI firmware security and secure booting without trusted hardware is more difficult to fix".
While the threat is still at the proof of concept stage, it is, according to Hudson, a significant threat. As it's a new exploit there are no methods for detection, it can spread easily, is persistent and stealthy.
While the hack demonstrated by Hudson required physical access to a Mac for injection, a more worrying issue was noted by Hudson. He suggested that the "Dark Jedi Sleep" attack, described by security researchers Rafal Wojtczuk and Corey Kallenberg at the Chaos Computer Club conference could be used to inject a modified version of Thunderstrike over the air when a Mac is waking up from sleep mode.
Interestingly, Apple could have shut this threat vector down had they retained trusted boot hardware in their systems. They did, at one time, have hardware for this but it was not used and was, eventually, removed by Apple from their hardware.
This article is brought to you by Enex TestLab, content directors for CSO Australia.
Upcoming IT Security Events
Feb 3rd, Feb 4th, Feb 6th 2015
Join @NirZuk #PaloAltoNetworks for Breakfast (lunch in Auckland) on keeping your enterprise safe from risk. Cyber attacks continue to increase in volume and sophistication leaving traditional security practices completely ineffective.
March 3rd, March 5th, March 9th 2015
Join CSO for the day@#csoperspectives and hear from @kimzetter @frankheidt @simplenomad
3 International Keynote speakers, 36 Key IT Security Industry Speaker, 21 Exhibitors, Security Analysts and many more.. Register today
Dont miss one of the biggest IT Security events in ANZ (registration is free, but seats are limited)