Slashdot Mirror
Schneier's Keynote At Linux.conf.au
Stony Stevenson writes "Computer security expert Bruce Schneier took a swipe at a number of sacred cows of security including RFID tags, national ID cards, and public CCTV security cameras in his keynote address to Linux.conf.au (currently being held in Melbourne, Australia). These technologies were all examples of security products tailored to provide the perception of security rather than tackling actual security risks, Schneier said. The discussion of public security — which has always been clouded by emotional decision making — has been railroaded by groups with vested interests such as security vendors and political groups, he claimed. 'For most of my career I would insult "security theater" and "snake oil" for being dumb. In fact, they're not dumb. As security designers we need to address both the feeling and the reality of security. We can't ignore one. It's not enough to make someone secure, that person needs to also realize they've been made secure. If no-one realizes it, no-one's going to buy it,' Schneier said."
CCTV almost never captures what you set out to catch. In many organizations, it's a knee-jerk reaction to some kind of incident. ie) Something got pinched, someone received an ass-kicking, etc. Even if you do catch it, you'll never be able to identify/recognize/charge/convict the person based on the video image alone. 4CIF at 30 fps is pretty much as good as it gets right now in most feasible installations. All you'll be able to say is, "Subject is hatless...REPEAT...HATLESS!" (And that's even if he's in the frame). The PTZ will just pan around aimlessly on a tour program, or be pointed at the wrong thing. However, wide-spread deployment of CCTV systems is still not futile; you just usually end up catching something that were never really looking for in the first place. People and vehicular traffic movements, facility useage, or realtime video of an incident in progress that just happens to be going-on in front of the lens. You can establish time frames of entry or exit, or use it to clue-you-in to the right path to finding the real evidence you're looking for. From a security systems perspective, more CCTV is better, but not to mitigate direct and specific threats. Only general ones. Or sometimes you just luck-out and with a good booby shot in the atrium of an office building.
For many of the same reasons there is no semblance of a secure electronic voting platform on the horizon. The reason is not that such a platform would be difficult to design. The reason is that it would not be profitable.
To be secure it would have to be open. In the case of voting platforms that means every line of code, every encryption algorithm, and all the hardware has to be open, published, and known. Nobody has yet figured out how to make enough money from such a system to outspend Diebold's lobbyists and earn considered from election officials.
No. What Bruce has realized is that, in the boardroom and the lunchroom (where almost nobody knows any better), security theatre often will kick the ass of real security practices because it's marketed by professional sales teams. It also often can be delivered for less (because it can be priced for what the market will bear).
If you want real security to be provided, you have to learn to sell it at least as well as the snake-oil. You have to make it sufficiently visible, but non-impeding, that people feel safe.
It's about understanding the human/political side of the equation that can make the difference between a successful deployment and a perceived failure.
Laissez lire, et laissez danser; ces deux amusements ne feront jamais de mal au monde. - Voltaire