Keeping Video Conferencing Security in Perspective …
The NSA is accused of spying on UN video conferences. What's the most likely scenario for what really happened?
Every few months the topic of video conferencing security rears its ugly head in the media. In the last week or so, the NSA has been accused of cracking the encryption guarding the United Nations' internal video conferencing systems. According to the well-known German magazine Der Spiegel, these NSA activities took place in the summer of 2012 and allowed the NSA to decrypt more than 450 communications. Apparently the Chinese tried to do the same thing, with unknown results.
First of all, the authors want to clarify what we don't know. We have no first-hand knowledge of any surveillance programs in use by any entity or agency, government or otherwise. We also don't have access to non-public details about the recent allegations. For example, we don't know which video calls were compromised or not compromised, which locations or people participated in these sessions, or what systems, services, or networks were in use.
Here's what we do know. Video conferencing can be very secure. The actual security level of a video call depends on a number of factors including (but not limited to) the equipment in use, the settings within the equipment, the networks in use, and the actual call settings.
Basically all video conferencing systems released in the last 10+ years have included 128-bit AES encryption. This means that calls between compatible systems should be encrypted using 128-bit encryption keys that are generated automatically at the start of each video session. Although AES is a commercial rather than military grade of encryption, AES packs quite a punch. According to a 2012 article in EE Times, a supercomputer would take 1 billion billion years to crack a 128-bit AES key using a brute force attack. Given that this is more than the age of the universe (13.75 billion years), this does not seem like a real-time risk.
And there is more. Despite the inherent security provided by 128-bit AES encryption, highly secure communication environments (e.g. the Department of Defense, Homeland Security, etc.) use additional methods to secure their communications. Most common are external, hardware-based encryption devices. These devices typically leverage multiple encryption keys including:
*A device-level physical encryption key that must be inserted in the device each time the system is used. These keys are typically exchanged every 30 days or more.
* Password keys that limit access to the device's functions and systems.
* Session keys that are automatically generated at the start of each session. In some cases, the session keys are even changed automatically during the session.
In addition, organizations conducting secure communications typically host their traffic on secure data networks that also use advanced encryption technologies.
So what does all of this mean? It is basically impossible to gain access to an encrypted video conference.
Now the question is--what really happened? Most likely, we will never know. What we do know is that most security issues we find in the video conferencing area are related to people not following basic security policies and procedures--either knowingly or unknowingly. Common causes of weak video conferencing security include:
* Turning off encryption on video systems
* Using outdated video systems that don't support encryption
* Failing to use the most current software on video systems/other devices
* Connecting to other devices (e.g. gateways, video bridges, etc.) that either don't support encryption or have encryption turned off
* Using software solutions or services that either don't encrypt or use less stringent encryption methods
* Failing to use proper passwords, not changing passwords often enough, or failing to keep those passwords secure.
These factors are NOT inherent weaknesses in video conferencing security; they are user-introduced weaknesses. This is analogous to putting money or jewels in a safe, and then either leaving the door wide open (on purpose for convenience or mistakenly) or writing the lock's combination on a post-it note hanging on the safe itself.
While not as likely, it is also possible that the UN's video systems themselves were compromised. For example, most video conferencing systems support remote monitoring and management. Some systems also support the streaming of video and audio content to remote users. These features are turned off by default and are password protected. If properly used, they are extremely valuable. However, if abused, these features can compromise security and provide access to secure information.
Finally, it is also possible (albeit unlikely) that someone gained access to the video systems and uploaded hacked firmware that provides back-door access to the audio and video content. (Remember the centrifuges in Iran!)
So given the limited information in hand, we believe it is far more likely that someone learned the IP address of one or more of the UN's video systems, and then snooped on non-encrypted conferences. Perhaps the video systems in some locations were not properly configured or had encryption disabled. Who knows?
But make no mistake--given sufficient time and resources, everything is "hackable." In the case of today's video conferencing systems, however, the chances of successfully eavesdropping on an encrypted media exchange are extremely small, even with the resources of the NSA.
Andrew Davis is co-founder and senior partner of Wainhouse Research, a boutique analyst firm that concentrates on collaboration products and services, including audio, video, and web conferencing and unified communications. Ira Weinstein is senior analyst and partner at Wainhouse.