Ch4Part4
Version 4, changed by sreents. 12/25/2005. Show version history
The main advantage in this system is its seamless verification. Unless you have your browser set to notify you of cookie exchanges, you can surf through sites that deposit and consume cookies with little interruption.
The disadvantage is the danger that your cookie file could be manipulated or even copied to other systems.1 It could also provide a key with which providers can learn about you. If a common cookie identifies you across a number of sites (because these sites have subscribed, for example, to a common tracking system), then, in principle, if you have revealed information about yourself in one of those places, the other places could know it as well.
Cookies are less secure than passwords, though of course they are doing something very different. This difference does not mean they are useless. The security we need is a function of the risk we are protecting against. Certainly, the NSA should not use cookies as the system for granting access to databanks of national secrets. But there is no reason why a portal site like Yahoo should not use cookies to figure out who you are and to give you the news that you have previously selected. You have little incentive to lie about this; Yahoo! has little reason to care if you do. With so little at stake, an unobtrusive if insecure system is perfectly adequate.
[UPDATE POSSIBILITY: the use of cookies across multiple sites, e.g., such as are used in various advertising networks or among affiliated sites. Cross-site cookies allow a Web site operator to "know" the identity not only of visitors who have registered on that particular site, but also on affiliated sites. Thus, while a system of universal digital identity has not developed, a system of federated digital identities has.]
A third technology would marry the benefits of the first two. This is the technology of digital signatures, which enables digital certificates,a kind of passport on the Internet. They would authenticate any type of information about a machine and, if unlocked by a pass-phrase or biometric device, about you—your name, your citizenship, your age, whether you are a lawyer. While the details of the architecture are many, suffice it for now to define them as encrypted digital objects that can be used to authenticate facts about someone. [UPDATE NEEDED: the use of a digital signature is now a common technique and not just a proposition for security and identity in cyberspace]
Digital certificates would reside on your computer (under at least some designs); a server would automatically (and invisibly) check the certificate as you entered the site. If you held the right certificate, you would be let in, and as you were let in, the server would then “know” the certified facts about you. It would “know,” that is, that you were a man, or that you came from Canada, or that you were over the age of twenty-five. And it could do or know all this without ever asking you anything at all. Certificates could become the kind of self-authenticating credential that we know in real space, but unlike in real space, there would be no limit to the facts the certificate could certify.
Digital certificates would make possible a secure system of identification that could operate as seamlessly as cookies but with much more data certified. They rely, however, on technologies of cryptography. To see how, we must take a detour into the workings of cryptography, and then consider again how this technology could be molded to the form of identification.
Cryptography: Confidentiality Versus Authentication
Here is something that will sound very extreme but is at most, I think, a slight exaggeration: encryption technologies are the most important technological breakthrough in the last one thousand years. No other technological discovery—from nuclear weapons (I hope) to the Internet—will have a more significant impact on social and political life. Cryptography will change everything.
I say this not because I have the space in this book, or ability in any case, to prove this claim to you. I say this to emphasize. It’s not important that you understand the underlying technologies, though it would be great if more people did. What is important is that you get a hint of the purposes to which these technologies can be turned, and the consequences of their power.
Cryptography is Janus-faced: it has an ambiguous relationship to freedom on the Internet. As Stewart Baker and Paul Hurst put it, cryptography “surely is the best of technologies and the worst of technologies. It will stop crimes and it will create new crimes. It will undermine dictatorships, and it will drive them to new excesses. It will make us all anonymous, and it will track our every transaction.”2
Footnotes
1 Any security problem comes not from the cookies directly (since they are simply passive data) but from applets (smaller applications) that might misuse the data collected. David Wille, in “Personal Jurisdiction and the Internet: Proposed Limits on State Jurisdiction over Data Communications in Tort Cases” (Kentucky Law Journal 87 [1999]: 95, 198–99), describes the security dangers with cookies. For a second description of the possible risks from cookies, see Jerry Kang, “Information Privacy in Cyberspace Transactions,” Stanford Law Review 50 (1998): 1193, 1227–29. Edit Delete
2 Stewart A. Baker and Paul R. Hurst, The Limits of Trust: Cryptography, Governments, and Electronic Commerce (Boston: Kluwer Law International, 1998), xv. Edit Delete
