BOSTON - That no one owns the Internet is taken as a truism. But the infrastructure on which the global network runs is owned by a handful of powerful corporations that can, and often do, use their control over the Internet backbone to their advantage in business negotiations. The influence these companies exert, some industry insiders fear, is strangling smaller companies and reducing customer choice.
"I can give you an example that shows that the Internet is owned by someone," says Jilani Zeribi, a senior analyst at market researcher Current Analysis in Sterling, Va. "Look at old peering arrangements, which were basically, 'I'll connect to your network, you carry my data and I'll carry yours.' The carriers started to realize that smaller ISPs were free-riding on their network, so they started charging for peering arrangements. Just the fact that someone wields that kind of power shows that someone owns the Internet."
A relatively small number of companies own and operate the fiber and cables that form the Internet infrastructure, with most of the power centered in the U.S., which has been the dominant nation in terms of Internet backbone and use. MCI WorldCom's UUNet division, ATT (T), GTE (GTK)'s Internetworking, Global Crossing (GX), Qwest Communications International (Q) and PSINet (PSIX) are among the U.S.-based major players. Globally, Telstra, the Australian national telecommunications carrier, and Global TeleSystems Group (GTS), which offers broadband in 20 European countries as well as various Asian incumbent telecommunications companies, have major ownership stakes.
Although industry executives, analysts and observers present the same general list of big infrastructure owners, they say there is no reliable way to measure exactly which company owns how much. One figure bandied about is that UUNet owns 30 percent of the backbone, with everyone else falling in line behind. AT&T claims the No. 2 spot.
Perhaps more important than placing numbers on ownership are the questions about what it means to have an ownership stake in the Internet, to have invested millions in getting the infrastructure up and keeping it running. What kind of power is conferred on the main Net owners, and how might they wield that power - for good or for nefarious ends?
Like analyst Zeribi, numerous others interviewed for this article pointed to peering arrangements as perhaps the prime example of how companies exert control. There also are issues related to co-location - where a number of ISPs, say, have their servers and other hardware located in the same spot, typically owned by a major vendor like UUNet - and to network access points, or NAPs, which are the spots where ISPs trade packets with other ISPs. NAPs are located globally, with some set up at college campuses and others in company buildings. Some NAPs are operated by organizations that charge no fees. Others are part of telecommunications vendors that charge monthly fees that can soar into thousands of dollars.
Working out co-location agreements can be tough because it involves "getting into the facilities of some of the big players," says Vince DiBiase, senior VP and chief sales officer for ICG Communications (ICGXQ), an integrated communications service provider based in Englewood, Colo. "It's not necessarily their fault," he adds. "At least on the surface it's not. They only planned so much space for the Internet and now it's so big."
The process can be "outrageously long and difficult and costly" to work through, he says.
Likewise, although NAPs are supposed to be neutral territory and therefore not subject to the same squabbles that result from working out peering arrangements between vendors, "getting in and out of the NAPs is tricky," DiBiase says.
While various people interviewed contend that the Internet business is so competitive that new players will always emerge to take over if the big guns become too powerful and wield their influence unfairly, DiBiase and others suggest that it's also the case that large infrastructure players
