LulzSec, Recent Hacks Show Government Agencies Unprepared
The U.S. Senate became the latest victim in a string of hacks into government and high-profile groups like the IMF and Lockheed Martin. Here’s what security experts say the Feds must do better.
By J. Nicholas Hoover, InformationWeek
LulzSec breached and released internal data from a U.S. Senate Web server Monday in the latest in a series of well-publicized attacks on high-profile government and government-related targets over the past several months that has also seen the compromise of the International Monetary Fund, Lockheed Martin, the Oak Ridge National Laboratory, the Group of 20, FBI partner InfraGard, Gmail accounts of government officials, and RSA’s SecurID authentication, which is used heavily in government.
While disparate actors, from so-called “hacktivist” groups like LulzSec and Anonymous out to exact revenge or embarrassment to well-organized nation states looking for government, economic, and military secrets, are likely behind the attacks, many of the attacks share in the fact that they have likely resulted from targeted spear phishing–carefully crafted emails and other messages designed to con unwilling recipients into installing malicious code. Experts say that no one cure fits all, and advocate a comprehensive defense. Government is making a big cyber push, but the repeated success of these attacks shows that government and organizations aren’t yet up to the challenge.
Symantec recently reported that targeted attacks are at a two-year high, but spear phishing is not a new tactic. Perhaps counter-intuitively, the fact that these attacks have become known may indicate that the government and other victims are at least getting better at uncovering and responding to the attacks, said Jim Lewis, senior fellow at the Center for Strategic and International Studies.
The federal government has made a major push for more than a year to overhaul agency use of the Federal Information Security Management Act, which regulates federal cybersecurity compliance. The government is moving FISMA from being a check-the-box auditing tool to an agent for the adoption of “continuous monitoring” of systems, vulnerabilities, threats, and attacks. New technologies like the Einstein intrusion detection and prevention systems are being installed across the government, and the administration’s latest FISMA reporting guidance focuses heavily on continuous monitoring.
And yet attacks are still getting through. “Most sophisticated organizations, their front end is hardened,” said Tom Kellerman, CTO of mobile security company AirPatrol and a member of the Obama administration’s cybersecurity commission who has done cybersecurity work for both the IMF and World Bank. “However, if you can compromise the credentials of the user, you can access all the things you want to access.”
SANS Institute director Alan Paller characterized continuous monitoring as only a sliver of the solution. There’s more work to be done, he said, in getting agencies to actually take action on the things they see as part of their monitoring, in tracking outgoing traffic from government computer networks, and especially in training and hiring talented security pros from a limited talent pool.
Kellerman said that there are a number of places where organizations are broadly failing in their defenses to targeted attacks, including a lack of two- and three-factor authentication that goes beyond just a password and something you know (like your mother’s maiden name), a failure of policy (like giving users system administrator privileges or allowing people to click links without checking their email headers for inconsistencies), and a lack of white listing certain email and browser behaviors (like failing to prohibit the download and install of unknown executables), especially on laptops and mobile devices.
“Apps should only be able to do and access certain things, only certain apps should be allowed to run, and you really need to have the capacity to limit the device when it’s doing sensitive things,” he said. “You need to be able to control, to be able to contextually manage the functionality of the device so you can’t exfiltrate data.” He said that the need was especially strong in the case of mobile devices and laptops that remotely access enterprise networks through VPNs.
Virualization is another possibly powerful preventative prescription. Security firm Invincea’s software, for example, isolates the Web browser in a virtual environment that’s instrumented to monitor for unexpected changes to the environment without needing to resort to a signature-based defense that’s often ineffective in today’s atmosphere of zero-day attacks.
Furthermore, some sort of liability should be imposed on employees who click on spoofed links or download poisoned files, Kellerman said. However, the defense will likely require more than just training. “We’re not going to train our way out of the problem,” said Anup Ghosh, CEO of cybersecurity firm Invincea and former cybersecurity researcher with the Defense Advanced Research Projects Agency.
Ghosh also recommends agencies and organizations place network breaks into their systems. “There’s no reason for the accounting subnet to be connected to the research and development subnet,” he said.
The highly publicized attacks may help force the hand of Congress, which has been pumping out cybersecurity bills but has thus far failed to get anything significant to the President’s desk. Many of those bills would require additional safeguards of government computers and those of critical infrastructure providers, which would likely include the likes of InfraGard, RSA, Lockheed, and Google.
“Incidents like this only reinforce the opportunity to do something,” Lewis said. “How many more of these do we need to see before we say as a country that we’re going to be serious on cybersecurity?” Lewis noted that while the Senate is beginning to move rapidly on cybersecurity, the House of Representatives is still bottled up with organizing and coordinating numerous “pieces of bills” that it has introduced.