(MIT Technology Review) The U.S. Internet’s infrastructure needs to be redesigned to allow the NSA to know instantly when overseas hackers might be attacking public or private infrastructure and computer networks, the agency’s leader, General Keith Alexander, said today.
Alexander spoke at the annual Def Con computer hacking conference in Las Vegas. It was a symbolic appearance that he said was motivated by a need to interest the hacker community in helping to make the Internet more secure.
Alexander, who is also commander of the U.S. Cyber Command, described the Internet as “at great risk from exploitation, disruption, and destruction.”
In recent years, many Internet users have become familiar with the idea that websites can be knocked offline by denial of service attacks, such as those employed by online activist groups such as Anonymous. “My concern is that it’s going to flow into destructive attacks that could have consequences for our critical national infrastructure and the Internet itself,” said Alexander.
The decentralized nature of the Internet, and the fact that the global network is built from a thicket of independent public and private networks, is limiting efforts to protect against such attacks, said Alexander, because it doesn’t allow the NSA or law enforcement to easily track Internet activity. “We do not sit around our country and look in; we have no idea if Wall Street is about to be attacked,” said Alexander.
The NSA is already running a trial with 17 U.S. defense companies intended to demonstrate technology that could be deployed to change that. Under the Defense Industrial Base (DIB) Cyber Pilot, Lockheed Martin and other companies set up their computer security systems to automatically alert the agency when the alarm is tripped. They automatically pass a summary of what was detected and the IP address associated with the event to the NSA over the Internet. “All you need to pass is the fact of a signature and IP address in real time, and we can take it from there,” said Alexander.