Efforts by the U.S. government to better identify cyberattackers will likely lead to violations of Internet users' privacy and anonymity, and technological means to attribute the source of the attacks may be inaccurate, privacy and cybersecurity experts said Thursday.
Witnesses at a U.S. House of Representatives subcommittee hearing disagreed about whether the government should explore new ways to attribute the sources of cyberattacks. Several cybersecurity experts have called for new attribution efforts, including trusted identification systems, but Robert Knake, an international affairs fellow for the Council on Foreign Relations, said oppressive governments would use new identification technologies to track their political enemies.
Proposals to label IP (Internet Protocol) packets with unique identifiers "would be far more useful for authoritarian regimes to monitor and control Internet use by their citizens than it would be in combating cyberwarfare, crime and nuisance behavior," Knake told the House Science and Technology Committee's subcommittee on technology and innovation.
For massive attacks, attribution of the attackers may not be difficult, because only a few nations have that capability, while low-level attacks don't rise to the level of national emergencies, he said. "In a lot of cases, we don't lack attribution, we lack response options," he added. "We don't know what we should do when we discover that the Chinese have hacked into Google."
Subcommittee Chairman David Wu, an Oregon Democrat, asked witnesses whether new ways to identify the source of cyberattacks would deter some attacks. In other types of conflicts the U.S. has faced, knowing the identity of potential attackers has kept both sides from aggression, he said.
Knake questioned whether the same would hold true for cyberattacks.
The Cold War theory of mutually assured destruction in nuclear attacks between the U.S. and the old Soviet Union is "unpersuasive" in a cybersecurity context because the U.S. is far more dependent on the Internet than many potential attackers, he said. The U.S. would have to respond with physical attacks to do the same amount of damage as the attacking nation did to it, he added.
Knake suggested that the U.S. government should focus more on preventing damage and protecting its systems than on attributing the source of attacks. But Ed Giorgio, president of cybersecurity vendor Ponte Technologies, called for new protocols that would identify users on sensitive networks. On less-sensitive parts of the Internet, people should have tokens, issued by a trusted third party, that establish their identity or tokens that give them anonymity, he said.
Attack attribution, while not effective today, is an "essential part" of the U.S. government's emergency response capabilities, Giorgio said.
"My comments are not focused on promoting what the ideal balance between privacy and security should be, but rather a challenge to those embracing the utopian view that both may be simultaneously within our grasp," he added.
But proposals to establish an Internet ID may not even be legal in the U.S., said Marc Rotenberg, president of the Electronic Privacy Information Center (EPIC). The U.S. has a long tradition of allowing residents to speak anonymously or pseudonymously, and courts have upheld that right, he said.
Such proposals would raise "significant" human rights and Internet freedom concerns, and they may not work, Rotenberg said.
"No matter how good attribution technologies are, attribution will probably still fail to identify the most sophisticated attackers," he said. "Because sophisticated attackers often obscure their trail by routing activities through multiple countries, complete attribution capability would require the implementation of coordinated policies on a near-impossible global scale."
Grant Gross covers technology and telecom policy in the U.S. government for The IDG News Service. Follow Grant on Twitter at GrantusG. Grant's e-mail address is firstname.lastname@example.org.