HTML 5's Privacy Problem

Lately there's been a lot of news and concern about perceived security and privacy problems in HTML 5. But while these concerns are certainly legitimate, for the most party there isn't really anything new here.The worries about security and privacy in HTML 5 started a few weeks ago, when Evercookie, a very difficult to remove tracking cookie, was released. Evercookie uses standard long-standing web capabilities, plus new data storage features in HTML 5, to store several versions of the cookie on a browser and also restore deleted copies of the cookie.

But the real noise started last weekend when the New York Times wrote an article about evercookie and the privacy problems of HTML 5. I'm sure pretty early this work week, many IT managers got concerned messages from higher-ups worried if the company was using this "dangerous new HTML 5 stuff".

So where are these problems in HTML 5 coming from? From one of the most promising and potentially useful new features of the web standard.

One of the new capabilities of HTML 5 is the ability to store lots of data locally. Among other things, this feature will make it possible to create web applications that run offline, without the need for third-party software or browser-specific features.

If you think about it, that's pretty cool. It would let you continue to use Google Apps or your favorite enterprise SaaS app or pretty much any web application even when you didn't have an Internet connection.

But there is of course that issue of letting web applications store data locally. Yes, this data could be useful content needed to let the application run efficiently and offline. But it can also be cookies and other nefarious tools used to track users without their knowledge or consent.

This is definitely a problem, and one that the W3C and the many companies that are involved in the development of HTML 5 will have to address. Luckily we still have probably a couple of years left until it is even an official W3C standard.

But as I said initially, this is hardly a new problem. After all, there are already many ways that a user can be tracked that are difficult to detect and stop.

In fact, just look at the Evercookie. Only a few of the ways that it hides itself have anything to do with HTML 5. The Evercookie also uses classic cookie storage, can hide itself in an image file and uses Flash cookies. Flash cookies in particular have been a privacy problem and for a while now web sites interested in quietly tracking users have taken advantage of them.

So yes, there are currently some clear privacy problems in the not-yet-a-standard HTML 5. Hopefully by the time it does become a standard, the W3C and the browser vendors will have come up with some ways to limit this problem.

But even if they do, don't for one second think that your privacy on the Web is protected. Even if you use every tool and technique available to block tracking, you can still be tracked on the web. For example, web bugs from major advertising, analytics and search firms are found on the majority of web sites and can be used to compile a profile of users even if they block all cookies.

So we all do what we can to protect our privacy as much as we can. But don't think HTML 5 is the main or only problem in privacy.