APT Attacks Can Be Launched Using Basic Skills: Imperva

Imperva Examines the “Non-Advanced Persistent Threat”

Not everything about the threats facing today’s enterprises is advanced.

In fact, as Imperva points out in a new report, the type of data breaches commonly associated with advanced persistent threats can sometimes be achieved without high levels of skill.

“Advanced Persistent Threat (APT) is a name given to attacks that specifically and persistently target an entity,” Imperva noted in a report entitled ‘The Non-Advanced Persistent Threat.’ “The security community views this type of attack as a complex, sophisticated cyber-attack that can last months or even years. The skill and scope required to instigate an attack of this magnitude and sophistication are believed to be beyond the reach of individual hackers.”

“Despite these common perceptions (see Wikipedia), our labs discovered that some techniques attributed to APT require only basic skills,” the report continued. “For example, there are simple ways to accumulate access privileges by attacking common Windows protocols. To provide evidence of this, the attacks we examined targeted commonly known, inherent weaknesses of the Microsoft NTLM protocol, and leveraged basic social engineering, Windows skills, and readily available software.”

“By attacking an organization using non-advanced techniques and still achieving the same goals, we learn that the problem of data exfiltration is in fact bigger than just the APT problem, and that companies should take a more holistic approach to protect their data centers and the data that resides in them,” blogged Barry Shteiman, director of security strategy at Imperva. “If a hacker can achieve the same results with or without APT, it becomes a different ball game.”

The report describes security flaws and attacks dealing with in NTLM, including pass the hash and NTLM relay attacks. According to the report, NTLM’s authentication response calculation does not require the plaintext password – it requires the NT or LM hash, which makes the LM or NT hash comparable to plaintext passwords. It also describes the NTLMv1 challenge response algorithm as weak, and states that an attacker who has access to the challenge and the response can calculate the LM or NT hash used for authentication. Given that the hashes are password equivalents, an attacker can authenticate as the user with the hashes.

Once an attacker compromises one machine, he or she gains access privileges of the current logged on user. The goal from then on is often to extend access by compromising more accounts. 

“NTLM protocol weaknesses provide an attacker a perfect opportunity to extend his access privileges to targeted resources – as long as those resources support NTLM authentication,” according to the report. “Note that Windows file shares and some databases – mainly MS SQL [Microsoft SQL Server] and Oracle – support windows based authentication using NTLM.”

The report also describes how attackers go about “poisoning the well” by introducing content to a shared folder that forces SMB traffic from users who browse that file to communicate with a compromised machine. 

“When planning a corporate security strategy, one of the things I truly believe is that the CIO should not only try to prevent hackers from getting in, but should assume that hackers can get in and that they may access your data,” Shteiman blogged. “From that point the game is all about knowing that this happens. Monitoring, audit, and blocking controls around data are essential in order to prevent the next big data breach.”