We've sacrificed a lot of privacy in the last few years in the name of security. Are we actually safer as a result?
The whole security vs. privacy dichotomy is a false one. There are many security measuresdoor locks, burglar alarms, tall chain-link fencesthat have nothing to do with privacy. Its only identity-based security that affects privacy, and there are limits to that approach. Ive repeatedly said that exactly two things have made airplane travel safer since 9/11: reinforcing cockpit doors and convincing passengers that they need to fight back. Those two things have no effect on privacy. Security measures that affect privacy, like ID checks, havent made us any safer. The real dichotomy is liberty vs. control. And real security comes from liberty plus privacy.
Protecting computer security is usually seen as a technological challenge, but you refer to it as an economic problem. Why so?
Because if you dont get the economic incentives right, no amount of technology will help. Security is a trade-off, and people will weigh the cost of security against the benefits. Its easiest to see this in a business environmentfor example, is an anti-fraud measure more or less expensive than the fraud it will preventbut its true everywhere: personally through nationally. These trade-offs arent made in some abstract greatest good sort of way; theyre made by people based on their own personal situation. And if the costs and benefits arent aligned, people wont make good trade-offs.
An example might make this clearer. A lot of identity theft comes from corporations not securing their databases filled with personal information. Of course, they could spend more money to increase security, but the economic incentives arent aligned: the risk of identity theft is borne by those people in the databases, not by the company. So it doesnt matter what kind of technologies you invent; it wont be worth it for the company to implement them. The way you fix this is by fixing the economics: making these data breaches costly to the company.
I found your writings about the psychology of security to be particularly interesting: about how we may feel we're secure when we're not, and vice versa. How does this gap affect our real world efforts to guard our security?
We end up with a lot of security measures that make us feel more secure, regardless of whether they actually make us more secure. This effect is most pronounced when its hard to evaluate the actual effectiveness of a security measure. Crime prevention measures are relatively easy to evaluate, because you can watch the crime rate go up or down. On the other hand, anti-terrorism measures can be very hard to evaluate, because there simply arent enough events to get a sufficient data sample. Fears, folk beliefs, and preconceived notions also make it hard to notice when the feeling of security doesnt match the reality. So we end up with a lot of security theater.
The big question: Our personal PC security. When people ask youas they often dowhat they can do to protect their PCs, you've been known to answer "nothingyou're screwed." But you readily admit the reality is more complicated. What are the most essential things people need to do?
Backup. Backup, backup, backup. For most people, the biggest security risk is losing their data. A regular backup will go a long way to making their computer more secure. And be sure to test those backups; theyre no good if the restore doesnt work. After that, invest in an anti-virus program and keep your patches up to date. Everything else is in the margins.