So-called “smart” locks and alarms are proliferating across people’s homes, even though hackers have shown various weaknesses in their designs that contradict their claims to being secure.
Now benevolent hackers in the U.K. have shown just how quick and easy it is to pop open a door with an attack on one of those keyless connected locks. And, what’s more, the five-year-old flaw lies in software that’s been shipped to more than 100 million devices that are supposed to make the home smarter and more secure. Doorbells, bulbs and house alarms are amongst the myriad products from 2,400 different vendors shipping products with the flawed code. Tens of millions of smart home devices are now vulnerable to hacks that could lead to break-ins or a digital haunting, the researchers warned.
For their exploits, the researchers – Ken Munro and Andrew Tierney from Pen Test Partners – focused on the Conexis L1 Smart Door Lock, the $360 flagship product of British company Yale. As relayed to Forbes ahead of the researchers’ report, Munro and Tierney found a vulnerability in an underlying standard used by the device to handle communications between the lock and the paired device that controls the system. The flaw meant the communications could be intercepted and manipulated to make it easy for someone in the local area to steal keys and unlock the door.
The problematic standard was the Z-Wave S2. It provides a way for smart home equipment to communicate wirelessly and is an update from an old protocol, Z-Wave S0, that was vulnerable to exploits that could quickly grab those crucial keys. Indeed, they were “trivial” to decrypt, according to Pen Test Partners’ research.
Z-Wave S2 is more secure than S0. It comes with a method for sharing keys known as the Diffie-Helmann exchange; it’s a highly-regarded, tested method for ensuring that the devices shifting keys between one another are legitimate and trusted. But whilst the Yale device, purchased by Munro and Tierney just a couple of weeks ago and kept up to date, used that S2 protocol, the researchers found it was possible to quickly downgrade the device to the older, much less secure key-sharing mechanism.
During the period when a user paired their controller (such as a smartphone or smart home hub) with the device, Munro and Tierney could ensure the less-secure S0 method was used. From there, they could crack the keys and get permanent access to the Yale lock and therefore whatever building it was protecting, all without the real user’s knowledge. They believe they could carry out their attack, dubbed Z-Shave, from up to 100 meters away.
“It’s not difficult to exploit,” Munro said. “Software Defined Radio tools and a free software Z-Wave controller are all that’s needed.” In 2016, hackers created a free program designed to exploit Z-Wave devices called EZ-Wave.
Yale owner ASSA ABLOY said it understood the Z-Wave Alliance was conducting an investigation into the matter and was in close contact. ASSA ABLOY will also be conducting its own investigation, a spokesperson said, adding that it was “constantly updating and reviewing products in line with the latest technologies, standards and threats.”
Munro told Forbes it should be possible to update many Z-Wave-based devices with a wireless update of both the app and the device. “However, it’s an issue with the Z-Wave standard, so would require a massive change by the Alliance, then an update pushed to all devices that support S2, which would likely stop them working with S0 controllers. And there are hardly any S2 controllers on the market. None in the U.K.,” he added.
Silicon Labs (SiLabs), the $4.5 billion market cap firm that owns the Z-Wave tech, admitted “a known device pairing vulnerability” existed. But it didn’t specify any upcoming updates and downplayed the severity of the attack, adding “there have been no known real-world exploits to report.”
The company referred Forbes to the first description of the S0 decryption attack, revealed way back in 2013 by SensePost, which determined the hack wasn’t “interesting” because it was limited to the timeframe of the pairing process. As a result, SiLabs said it didn’t see the S0 device pairing issue “as a serious threat in the real world” as “there is an extremely small window in which anyone could exploit the issue” during the pairing process, adding that a warning will come up if a downgrade attack happens. “S2 is the best-in-class standard for security in the smart home today, with no known vulnerabilities,” the spokesperson added, before pointing to a blog released by SiLabs Wednesday.
Munro said it would be possible to set up an automated attack that would make it more reliable. “It should be easy to set up an automated listener waiting for the pairing, then automatically grab the key,” he said.
The company said the problem existed because of a need to provide backwards compatibility, as a spokesperson explained: “The feature of S2 in question – device pairing – requires both devices have S2 to work at that level. But of course the adoption of this framework across the entire ecosystem doesn’t happen overnight. In the meantime, we do provide the end user with a warning from the controller or hub if an S0 device is on the network or if the network link has degraded to S0.”
Munro was flabbergasted at the vendor’s overall response. “After attempting responsible disclosure and getting little meaningful response, on full disclosure Z-Wave finally acknowledge that it’s been a known issue for the last few years. Internet of Things (IoT) devices are at their most vulnerable during initial set-up. S2 Security does little to solve that problem.”