Thursday, August 12, 2010

Upping the Ante: A Thought Exercise

WARNING: This post contains a trip into the wondrous what-if-land to explore a different means of increasing security for the end-user.

I was thinking the other day about how the landing pages for browsers (and other products) usually tout security as one of its main attributes. Below are a few examples:

What makes Firefox the best?

Keeping you safe while you surf is our top priority, which is why Firefox includes advanced anti-phishing and anti-malware technologies plus features like private browsing and “forget this site” to ensure your privacy.

Plus, our open source security process means we have experts around the globe working around the clock to keep you (and your personal information) safe.

Google Chrome:
Google Chrome includes features to help protect you and your computer from malicious websites as you browse the web. Chrome uses technologies such as Safe Browsing, sandboxing, and auto-updates to help protect you against phishing and malware attacks.

Internet Explorer:
Browser Security

Other browsers leave you more exposed to phishing, malware and other online threats. Or they require you to download and configure third-party add-ons just to get the security Internet Explorer 8 comes with right out of the box

Safe and secure

Reduce your exposure to threats on the Web. Industry-leading security and features such as Web Threat Protection and a security bar, ensure your safety.

To me each of them is being _very_ conservative when it comes to describing how secure their product is. Each of those companies has extremely intelligent people from the security field working for them, about whom each company could boast. They could also talk about how few bugs have been found in their product, or their awesome fuzzing-farm that should dramatically reduce the number of 0days available for researchers to find. Or, if they wanted to get really crazy, they could even publicly bash their competitors, saying things like "Browser A sucks! BB security vulnerabilities were found in it in the past CC months! Only DD vulns were found in ours!"

What if companies actually _did_ focus on publicly bashing their competitors? The end goal for a company making a product is to increase their user-base and hence their profits, right? If they could make their competitor's product look like swiss cheese because of the number of security holes in it, they might very effectively keep existing users from switching to a competitor. They could also try to "prove" how little their product resembles swiss cheese, drawing in more users.

So far, this all doesn't seem too far-fetched. Often, I think of competing companies as being relatively nice towards each other when they find security vulnerabilities in the competitor's product. The vuln might get reported responsibly and then isn't really mentioned once it's fixed. But this is all different in what-if land. Imagine a world where they weren't publicly "nice" to each other. In this world, each company's security team would have two main job functions: fix our product, and find security holes in the competition so we can publicly humiliate them!

If this were the case, everything would be nuts and highly entertaining! Security marketing for a product would be more along the lines of "Don't use product EE! Our security team alone has recently found FF vulnerabilities in their product, we have only had [some # smaller than FF]!" or "Our security team finds more bugs in our competitor's products than they can manage to find and fix on their own! Imagine how secure our product is with such a team securing it!" or even more crass "Product GG has been found to be vulnerable to HH! We're not, use us! If you use them, we'll pwn you with the 0days we've been saving!" Once a company released a statement publicly bashing a competitor, the competitor would be forced to defend themselves in some form or another in order to keep their existing user-base. They might counter-attack, releasing advisories and embarrassing the attacker, or they might show how security feature MM effectively thwarts those attack vectors. If a company has no comeback, their user-base will dwindle until they can save face before their users.

Mainstream media would, of course, have a heyday with any and all public bashing. News stories with headlines such as "Company II publicly humiliates Company JJ! JJ declares war on II!" would abound, increasing the drama and upping the ante and humiliation.

But who would all of this actually benefit? Well, that should be fairly obvious, actually: the end-user! (Hey, isn't that what all this security stuff is about anyways?!) With all of the public bashing, humiliating, security dream-team putting-togethering going on, the actual products themselves will become more secure, either directly or indirectly. The rate at which bugs are found and fixed in a product would be amplified tremendously. New security features would constantly be thought up to stave off attacks and to undermine the competitor's claims. New policies would be put into place in companies to _quickly_ provide fixes for the end-user. Companies would scour the globe looking for the best in the security field to help them defend their product and attack the competitors. In the end, by the time the product reaches the end-user, it would be more secure. The end-user would also have a good time sitting back and enjoying the drama! Heck, companies could even sell tickets! We could make a sporting event out of it!

However, this wouldn't only help the end-user: it would help security researchers tremendously. As a result of the need for sufficiently good bash-ammunition and defensive armor a company will need against its competitors in order to stay alive, large companies that currently do not pay for vulnerabilities would be forced to pay. Not only would they be forced to pay, they would have to compete for security researcher's attention by constantly increasing the incentive. It's a win-win situation!

Of course there are some pitfalls in this thinking that I haven't mentioned, such as a large company picking on a small start-up who doesn't yet have the means to compete or defend themselves, or companies targeting specific individuals in the competition instead of the company they work for. In such a case, referees would have to be appointed who could rebuke the offending, rule-breaking company and publicly shame them for being a bully. The larger company then might have to help the smaller company out with it's security team and policies until it's able to defend itself. If an individual was personally harassed, the company should have to publicly apologize and send him a fruit-basket and a "Get-Well" card.

Crazy idea, huh? There are several other points I thought about bringing up, but I think I demonstrated what I wanted to. Would this really work? Probably not. It'd be crazy if it did though. Am I serious about this post? Partly (remember, this was only a thought-exercise!)

Even though this is an extreme example, the point I really wanted to make with this post is how different things might be for the end-user if companies personally had more at stake for creating (and selling) insecure products. My conclusion is that things could be very different from how they are today. Companies would have bigger/better security teams, different bug-fixing policies requiring fast turn around times for patch development, and the incentive for reporting vulnerabilities directly to the vendor would be at least better than the possibility of having your name in the acknowledgment section of an advisory. I think it might be a good thing.

PS - On a side note, in case you needed more food for thought: Does the problem of finding and fixing every security vulnerability in a product belong in P, NP, or is it NP complete?