Fears are rising that IP cameras can and willl be hacked. At Defcon, a demonstration showed an IP camera's feed intercepted and replaced by a fraudulent video, allowing a hypothetical suspect to steal an object right in front of the surveillance camera; thus bringing Hollywood to 'real life.'
What Do You Think?
Demo of the Hack
Here's a demo of the hack (the theft occurs at the end of the clip). Note the company that does the hack sells software to prevent it.
Bigger Risks Routinely Accepted
As titilating as this demo may be, there are far bigger risks that most real-world security organizations accept every day, such as:
Most security cameras are not watched live. For all those cameras, there's no need for any fancy hacks. Just walk on in. On the way out, find the recorder and take it with you.
When security cameras fail, almost no one responds immediately. At best, a trouble ticket or call is opened and the camera is checked in the next few business days. If the cameras are being monitored live, simply shut down the recorder or the power to the recorder/cameras. Most operations will see this as a nuisance but will not shut down the building (casinos, as always, the exception).
Difficulty to Do in a Real Environment
It's one thing to do this in a demo, it's far harder to pull this off in a real environment. Let's say you are one of the very few organizations who both watches cameras live and takes immediate action to cameras going out. The attacker would still need to:
Get access to the internal LAN of the target organization.
Pull this hack off against many cameras. These types of organizations are going to have dense camera coverage, which means 3, 5, 10 or more cameras need to be commandeered.
The attacker will also have to figure out where these cameras are - which generally is not easy. Steal the CAD drawings? Hack in to the VMS system to see the layout? Certainly theoretically possible but not easy to do.
The demo presumes the use of standard signaling protocols and CODECs. IP video surveillance is famous for its lack of standards. The attacker will have to know which proprietary interface each camera uses and have solutions for each variety. Good luck.
If the attacker is this cunning, intelligent and determined, wouldn't there be higher value targets? Steal critical information, access financial accounts, etc. Or do this legally by becoming a quant at an investment bank?
This type of attack is security theater - the type of risk that sounds exciting and threatening but is simultaneously unrealistic and ignores more fundamental risks that should be addressed. Maybe maximum level security operations should examine this but I suspect even they have more basic flaws in their video surveillance that need to be addressed first.
Xandem's "full coverage motion tracking technology" is unlike any intrusion technology we have seen.
We bought their new system and tested it...
The world's leading video surveillance information source, IPVM provides the best reporting, testing and training for 10,000+ members globally. Dedicated to independent and objective information, we uniquely refuse any and all advertisements, sponsorship and consulting from manufacturers.