The proposal is part of a final version of a report, "The National
Strategy to Secure Cyberspace," set for release early next year, according
to several people who have been briefed on the report. It is a component
of the effort to increase national security after the Sept. 11 attacks.
The President's Critical Infrastructure Protection Board is preparing
the report, and it is intended to create public and private cooperation to
regulate and defend the national computer networks, not only from everyday
hazards like viruses but also from terrorist attack. Ultimately the report
is intended to provide an Internet strategy for the new Department of
Such a proposal, which would be subject to Congressional and regulatory
approval, would be a technical challenge because the Internet has
thousands of independent service providers, from garage operations to
giant corporations like American Online, AT&T, Microsoft and Worldcom.
The report does not detail specific operational requirements, locations
for the centralized system or costs, people who were briefed on the
While the proposal is meant to gauge the overall state of the worldwide
network, some officials of Internet companies who have been briefed on the
proposal say they worry that such a system could be used to cross the
indistinct border between broad monitoring and wiretap.
Stewart Baker, a Washington lawyer who represents some of the nation's
largest Internet providers, said, "Internet service providers are
concerned about the privacy implications of this as well as liability,"
since providing access to live feeds of network activity could be
interpreted as a wiretap or as the "pen register" and "trap and trace"
systems used on phones without a judicial order.
Mr. Baker said the issue would need to be resolved before the proposal
could move forward.
Tiffany Olson, the deputy chief of staff for the President's Critical
Infrastructure Protection Board, said yesterday that the proposal, which
includes a national network operations center, was still in flux. She said
the proposed methods did not necessarily require gathering data that would
allow monitoring at an individual user level.
But the need for a large-scale operations center is real, Ms. Olson
said, because Internet service providers and security companies and other
online companies only have a view of the part of the Internet that is
under their control.
"We don't have anybody that is able to look at the entire picture," she
said. "When something is happening, we don't know it's happening until
it's too late."
The government report was first released in draft form in September,
and described the monitoring center, but it suggested it would likely be
controlled by industry. The current draft sets the stage for the
government to have a leadership role.
The new proposal is labeled in the report as an "early-warning center"
that the board says is required to offer early detection of Internet-based
attacks as well as defense against viruses and worms.
But Internet service providers argue that its data-monitoring functions
could be used to track the activities of individuals using the network.
An official with a major data services company who has been briefed on
several aspects of the government's plans said it was hard to see how such
capabilities could be provided to government without the potential for
real-time monitoring, even of individuals.
"Part of monitoring the Internet and doing real-time analysis is to be
able to track incidents while they are occurring," the official said.
The official compared the system to Carnivore, the Internet wiretap
system used by the F.B.I., saying: "Am I analogizing this to Carnivore?
Absolutely. But in fact, it's 10 times worse. Carnivore was working on
much smaller feeds and could not scale. This is looking at the whole
One former federal Internet security official cautioned against drawing
conclusions from the information that is available so far about the
Securing Cyberspace report's conclusions.
Michael Vatis, the founding director of the National Critical
Infrastructure Protection Center and now the director of the Institute for
Security Technology Studies at Dartmouth, said it was common for proposals
to be cast in the worst possible light before anything is actually known
about the technology that will be used or the legal framework within which
it will function.
"You get a firestorm created before anybody knows what, concretely, is
being proposed," Mr. Vatis said.
A technology that is deployed without the proper legal controls "could
be used to violate privacy," he said, and should be considered carefully.
But at the other end of the spectrum of reaction, Mr. Vatis warned,
"You end up without technology that could be very useful to combat
terrorism, information warfare or some other harmful act."