Advanced persistent threats -- a term that's become much maligned since the media locked onto it -- describes attackers that are targeting specific companies and data, rather than searching for vulnerable targets of opportunity. Persistent attackers stole oil field exploration data from ExxonMobil, information on the Joint Strike Fighter from Lockheed Martin and Northrup Grumman, and sensitive data on SecurID tokens from RSA. For many in the industry, the question is no longer if they have been breached, but how deeply, says Richard Bejtlich, chief security officer of Mandiant.
"No one has been able to stop these guys, no one," he says. "They remain a problem for every company with valuable intellectual property."
Separating persistent threats from more opportunistic cybercrime-focused attacks is not easy, but can help inform defense, according to security experts. Block an opportunistic attack and the crisis is averted; block a persistent attacker and they will come back tomorrow, says Toralv Dirro, security strategist for McAfee's Labs in the Europe, Middle Eeast and Africa region.
"If someone is a victim of a targeted attack, there are patterns," Dirro says. "They should really follow up on identifying those patterns."
In many cases, the patterns are not clear. Even "advanced" attackers will only use, for example, the minimum force necessary to compromise a network. In some cases, attackers have rented botnets; in others, they've used standard cybercrime tools.
"It is never a case of, oh, they are using Poison Ivy, so it's APT -- everyone is using Poison Ivy," Mandiant's Bejtlich says. "It really comes down to a lot of analysis to figure out what is going on."
Attacks that link back to specific nations, especially China, could indicate the attack qualifies as an advanced persistent threat, but tracking the origin of attacks is notoriously unreliable. While the United States hosts the lion's share of command-and-control servers used for malicious attack -- both cybercrime and espionage -- most of the attacks are coming from China, says Bejtlich. The company tracks 12 Chinese groups responsible for much of the advanced persistent threat (APT) related attacks.
"The amount of activity you get from all the other people out there is swamped by the stuff coming out of China," he says.
Generally, however, there are some common components, say other security professionals. Advanced persistent threats typically are attacks that use an employee as a beachhead into the network, and then move laterally from computer to computer. Malicious programs that infiltrate a network are likely a sign that the attack is more than mere crimeware, says Anup Ghosh, CEO of browser security firm Invincea.
"They will move laterally, and that is a distinguishing feature," says Ghosh. "They will do internal recon of machines, where a typical banking Trojan will not."
Identifying the entry point -- where an attacker got into a company's network -- is a key aspect of identifying and responding to an advanced attack, says Sean Brady, director in RSA's identity and data protection group. Watching for the telltale signs of reconnaissance is almost impossible, because so much reconnaissance is done on social networks, Internet databases and other sources outside the control of company.
"Think about information that LinkedIn supplies," he says. "You can map a company just by information gained from LinkedIn, and that is not a knock against LinkedIn, that's just the information they provide as a social network."
Instead, companies should focus on employees, Brady says.
The attack against security giant RSA earlier this year, for example, started with two e-mail messages that appeared to carry an attached recruitment plan and ended by reportedly compromising the company's database of cryptographic seeds for its SecurID tokens. The targets of the e-mails were not executives but low-level employees, a tactic that shows that companies have to worry about protecting all potential entry points into their network, RSA stated in a report from an advanced threats working group that convened in July.
"The advanced threat is aimed at a specific targeted point of entry," says Brady."And targeting a specific individual, and the person is almost never the ultimate target."
Educating users and focusing on securing the end user can help both detect and mitigate persistent attacks. An industry confab held by RSA and TechAmerica created a roadmap for dealing with advanced persistent threats. Industry has to get used to a state of compromise, concluded the group in a report released in September. Because attacks have moved away from technology and moved to targeting people, companies that have a good awareness of the security situation will be able to counter advanced threats best.
"Remember, an advance threats always starts with an individual, with people," says Brady. "The best approach is to narrow the impact, and then work on a response plan to keep out the attacks."
Have a comment on this story? Please click "Discuss" below. If you'd like to contact Dark Reading's editors directly, send us a message.