Sunday, February 15, 2009 - The Progress & Freedom Foundation Blog

Internet Security Concerns, Online Anonymity, and Splinternets

What would it take to create a more secure Internet? That's what John Markoff explores in his latest New York Times article, "Do We Need a New Internet?" Echoing some of the same fears Jonathan Zittrain articulates in his new book The Future of the Internet, Markoff wonders if online viruses and other forms of malware have gotten so out-of-control that extreme measures may be necessary to save the Net. Compared to when cyber-security attacks first started growing over 20 years ago, Markoff argues that:

[T]hings have gotten much, much worse. Bad enough that there is a growing belief among engineers and security experts that Internet security and privacy have become so maddeningly elusive that the only way to fix the problem is to start over.

Like many others, Markoff fingers anonymity as one potential culprit:
The Internet's current design virtually guarantees anonymity to its users. (As a New Yorker cartoon noted some years ago, "On the Internet, nobody knows that you're a dog.") But that anonymity is now the most vexing challenge for law enforcement. An Internet attacker can route a connection through many countries to hide his location, which may be from an account in an Internet cafe purchased with a stolen credit card. "As soon as you start dealing with the public Internet, the whole notion of trust becomes a quagmire," said Stefan Savage, an expert on computer security at the University of California, San Diego.

Consequently, Markoff suggests that:
A more secure network is one that would almost certainly offer less anonymity and privacy. That is likely to be the great tradeoff for the designers of the next Internet. One idea, for example, would be to require the equivalent of drivers' licenses to permit someone to connect to a public computer network. But that runs against the deeply held libertarian ethos of the Internet.

Indeed, not only does it run counter to the ethos of the Net, but as Markoff rightly notes, "Proving identity is likely to remain remarkably difficult in a world where it is trivial to take over someone's computer from half a world away and operate it as your own. As long as that remains true, building a completely trustable system will remain virtually impossible." I've spent a lot of time writing about that fact here and won't belabor the point other than to say that efforts to eliminate anonymity for the entire Internet would prove extraordinarily intrusive and destructive -- of both the Internet's current architecture and the rights of its users. There's just something about a "show-us-you-papers," national ID card-esque system of online identification that creeps most of us out. That's why I spend so much time fighting age verification mandates for social networking sites and other websites; it's the first step down a very dangerous road.

But what if we could apply such solutions in a narrower sense? That is, could we create more secure communities within the overarching Internet superstructure that might provide greater security? Markoff starts thinking along those lines when he suggests...

What a new Internet might look like is still widely debated, but one alternative would, in effect, create a "gated community" where users would give up their anonymity and certain freedoms in return for safety.
... but he is still thinking in terms of a replacement model for the entire Internet, which would be misguided for the reasons I stated above. We don't want to force a single, intrusive, anonymity-killing replacement model on the entire online universe. Starting over isn't even possible in a practical sense.

It's a shame that Markoff didn't interview my old colleague Wayne Crews for his story because Wayne has outlined an alternative framework worth considering. For many years, Wayne has been preaching about "spinternets," or the notion that we need to start thinking about how develop not just one better Internet, but many better Internets. In a visionary piece for Forbes back in early 2001, Wayne argued that the solution to the growth of various online concerns "is more Internets, not more regulations":

The Internet needs borders beyond which users can escape damaging political resolutions of these battles, which are rooted in the Internet's nonowned, common-property status. Conflicting legislative visions in a cyberspace populated by exhibitionists at one extreme and would-be inhabitants of gated communities on the other, reveal the basic truth that not everybody wants or needs to be connected to everybody else.

Again, there's that notion of "gated communities" that Markoff brought up. It's not for everybody, but those seeking greater security could perhaps find it inside such online communities. Of course, others who wanted a different experience could start a completely different gated community under Wayne's model.

But the problem with this notion, quite obviously, is that very few people want to stay inside their gated communities all the time. In the physical world of gated communities, for example, members of it still like to get out of there once and awhile to visit shops, events, parks, friends and family, etc. The same goes for the Internet. Just ask all those former denizens of AOL's gated community. For awhile, many of them -- over 25 million strong at the zenith of its popularity -- were content to spend most of their digital day inside the walls of Case's Castle. Gradually, however, they felt the need to explore outside those walls. And so they did. A mass exodus ensued and the walls came crumbling down around AOL's gated community.

But that doesn't necessarily mean the idea of online gated communities is entirely dead. There are certainly many closed, tightly-controlled networks out there already -- mostly in corporate or government environments -- that offer a glimpse of how such a model might work in practice. Also, smaller social networking sites aimed at kids provide another example since they are usually tightly-controlled walled gardens that offer much greater security.

But Wayne was always thinking of something bigger -- much bigger -- than just closed corporate / government networks. He was thinking about a world of many different Internets that didn't necessarily have a back door to the broader Internet. Think of it as many parallel, but unconnected digital systems and networks, each serving a different set of values and cultures with unique rules.

Wayne envisioned the primary critique of this model in his original piece, noting that "it will be criticized as Balkanization." Indeed, Sonia Arrison called it "techno-isolationism, which goes against the very spirit that makes the Internet great." Indeed, it certainly would destroy something very precious about the current Internet -- universal connectivity and openness. But that's sort of the point, isn't it! Universal connectivity and openness have given us many wonderful things, but some troubling things, too. That's what Markoff was getting at in his NYT piece, and it's part of what Wayne was aiming to address with his splinternets idea.

But do we really want to encourage a world of multiple Internets where, presumably, they are split right down to the root? In other words, there wouldn't be a common language for networks to communicate or a way to access many sites and services outside the particular Net you are on at any given time. It would be the equivalent of living on different digital planets that never linked or communicated.

I think it's unlikely we'll ever get there, and if we did it would likely be driven by global governments challenging ICANN and existing Internet governance structures. In other words, the DNS root would be completely split by some countries (China?) who didn't want to play by the same rules as the rest of the interconnected world, or who wanted to try to impose a different vision upon a new, competing global network.

But might there be a way to find a happy middle ground between the Wild West commons of the current Net and the "techno-isolationism" of Wayne's splinternet model? Perhaps "Splinternet-lite" is the solution. Within the confines of the existing Internet superstructure, there are ways to create walled gardens today and limit the number of back doors to the broader Net. Again, the smaller social networking sites and virtual worlds aimed at kids already do that. Once you're in there, you're in a very different world. You have to be fully verified before you're even let in the door, and once you're inside their are tight limits on what you say, do, and explore. And you'll get booted out pretty quickly if you break the rules. The result is greater safety and peace-of-mind for kids and parents alike. It's a less clear, however, how that model would "scale up" and apply to the entire universe of online networks. I think we'll have to be content with small patches of security within a world of insecurity. That's the cost of the openness and interconnectivity that the Net current gives us.

In sum, there is no clear answer to John Markoff's question, "Do we need a new Internet?" We certainly could do more to address the problems with the current Net, but upending it and starting over isn't likely an option. More micro-splinternets within the overarching Net superstructure, however, might help those who are particularly risk-conscious find safe haven from various cyber-security fears. But it won't shelter them from those problems completely.

posted by Adam Thierer @ 4:43 PM | Free Speech , Innovation , Internet Governance