More EAS devices vulnerable now than when vendors were alerted in January

Gunter Ollmann, CTO, Security, Microsoft Cloud and AI Division

July 15, 2013

3 Min Read

Every month or so in the U.S. you'll likely hear the "BARRP BARRP BARRP ... This is a Test Message" over the radio as the station you're listening to conducts its mandatory testing of the Emergency Alerting System (EAS). This is the same system you'll have also seen in many alien invasion and impending apocalypse movies where the president of the United States interrupts all TV broadcasts with an important message.

It's also the same system that was briefly hacked in Montana, Michigan, and New Mexico in February to warn local residents of zombie attacks.

That particular hack occurred a month after the vendor for the system had been alerted to a bunch of critical vulnerabilities uncovered by Mike Davis of IOActive Inc., and several months prior to the firmware updates being available for owners of the system to apply.

As it happens, the vulnerability exploited in the February "zombie attack" alerts appears to have been as trivial as knowledge of default shipping passwords that hadn't been changed by the TV stations.

Last week CERT and IOActive(PDF) released advisories detailing the nature of the vulnerabilities and provided links to the updated firmware patches by the vulnerable vendors -- Digital Alert Systems and Monroe Electronics.

There was some confusion at the time because the vulnerable vendors appeared to have released patches for some of the undisclosed vulnerabilities earlier; their press release of June 13 is dated April 24th (the date of when the vendors supposedly began outreach to their vulnerable customers).

Regardless, the most critical vulnerabilities -- the compromised SSH root key, default passwords, and predictable password generation -- allow attackers to trivially take control of the vulnerable systems and override station broadcasts.

The EAS itself is categorized as critical national infrastructure, yet it appears to still be largely vulnerable to attack even months after various security updates and alerts have been released. In a blog late last week, the original discoverer of the vulnerabilities, IOActive's Davis, indicated that more of the system is vulnerable to attackthan it was when he'd alerted the vendors back in January.

I've heard a few ill-informed rumblings that even if the EAS were hacked again, it would just be a nuisance. In the most trivial case, if the system were to be hacked again and the hackers were to issue repeated "zombie apocalypse" warnings, people would likely lose confidence and trust in the system -- meaning that in the time of a real emergency, people may not follow the real advice, and lives could be lost.

Alternatively, if I were employed by a foreign military organization, hacking the EAS would be fairly high up on my list of critical infrastructure assets to target. Not only could I usurp control of the system to release disinformation and sow confusion, I could probably do so without falling afoul of too many articles of war. In addition, with access to the system I could also deny broadcasts by the legitimate broadcasters -- either by continually broadcasting my own messages or by replacing SSH keys and changing passwords to prevent their future access.

To prevent these scenarios from coming into play, it is critically important that the various broadcasters and their system administrators secure the Digital Alert Systems and Monroe Electronics EAS devices --- by applying the latest security updates, changing default passwords, and doing their best to restrict Internet access to the devices from unauthorized IP addresses. At the present time, there are lots of vulnerable devices and knowledge of the vulnerabilities is now public.

BARRP BARRP BARRP ... This is Not a Test!

Gunter Ollmann, CTO, IOActive Inc.

About the Author(s)

Gunter Ollmann

CTO, Security, Microsoft Cloud and AI Division

Gunter Ollmann serves as CTO for security and helps drive the cross-pillar strategy for the cloud and AI security groups at Microsoft. He has over three decades of information security experience in an array of cyber security consulting and research roles. Before to joining Microsoft, Gunter served as chief security officer at Vectra AI, driving new research and innovation into machine learning and AI-based threat detection of insider threats. Prior to Vectra AI, he served as CTO of domain services at NCC Group, where he drove the company's generic Top Level Domain (gTLD) program. He was also CTO at security consulting firm IOActive, CTO and vice president of research at Damballa, chief security strategist at IBM, and built and led well-known and respected security research groups around the world, such as X-Force. Gunter is a widely respected authority on security issues and technologies and has researched, written and published hundreds of technical papers and bylined articles.

Originally, Gunter had wanted to be an architect but he lost interest after designing retaining walls during a three-month internship. After that, he qualified as a meteorologist, but was lured to the dark side of forecasting Internet threats and cyberattacks. His ability to see dead people stoked an interest in history and first-millennium archaeology.

Keep up with the latest cybersecurity threats, newly discovered vulnerabilities, data breach information, and emerging trends. Delivered daily or weekly right to your email inbox.

You May Also Like


More Insights