The U.S. Department of Defense did not receive personal data on users of Internet privacy service Tor through a government-funded project to detect vulnerabilities, a Defense spokeswoman told Reuters on Tuesday, 5 August 2014.
“This particular project was focused on identifying vulnerabilities in Tor, not to collect data that would reveal personal identities of users,” said Defense Department spokeswoman Lieutenant Colonel Valerie Henderson, adding that the National Security Agency also did not receive data.
The project was conducted by two researchers at Carnegie-Mellon University’s Software Engineering Institute with funding from the Defense Department.
She did not rule out the FBI or other agencies obtaining the data. The FBI and Carnegie-Mellon declined to comment.
Funded in large part by other arms of the government, Tor hides the Internet protocol addresses of users by routing their traffic through multiple layers of volunteered servers.
In a note last week on Tor’s website, Tor Project leader Roger Dingledine said the service had identified computers on its network that had been quietly altering Tor traffic for five months in an attempt to unmask users connecting to what are known as “hidden services,” which include drug bazaars and whistleblower sites.
Dingledine said it was likely the attacking computers, which were removed on July 4, had operated on behalf of the Software Engineering Institute team.
He warned then that “users who operated or accessed hidden services from early February through July 4 should assume they were affected.”
Dingledine said the physical locations where the hidden services were housed could have been exposed, although probably not which content was viewed by a visitor.
In an email to Reuters, Dingledine said that Carnegie-Mellon had stopped cooperating and would not share more information about the effort.
The researchers had planned to describe their work at the Black Hat security conference that begins Wednesday in Las Vegas but the university canceled the talk amid the controversy.