Attacks On Volatile Memory Can Be Detected, Researchers Say

In-memory attacks create processing delays that give hackers away, Triumfant research says

Tim Wilson, Editor in Chief, Dark Reading, Contributor

October 3, 2013

2 Min Read

Elusive attacks on a computer's volatile memory can be detected through a detailed analysis of processor behavior, according to new research.

Researchers at security vendor Triumfant have discovered that in-memory attacks create a significant delay in system calls that is typically beyond the normal variance of processing time. The ability to detect such attacks -- which have generally eluded most security tools because they attack data that is not stored -- could enable enterprises to interrupt the attacks before they can do any damage, Triumfant says.

"There's a temporal dimension to in-memory attacks that is detectable," says John Prisco, CEO of Triumfant. "We're seeing delays in system calls that are two or three times the norm, and it's possible to isolate those processes and shut them down."

In-memory attacks, recently referred to as Advanced volatile threats (AVTs), enable an attacker to access a computer's random access memory (RAM) or other volatile memory processes to redirect a computer's behavior. AVTs allow attackers to steal data or insert malware, but because they are never stored in long-term memory, they can be difficult to detect.

Industry experts suspect that in-memory attacks are on the increase because they evade the prevalent defenses that rely on attack signatures and malware behavior analysis. Oded Horovitz, CEO and founder of security firm PrivateCore, last month presented his company's findings on server in-memory attacks (PDF) and recommended tools for encrypting such data.

"Hacking hasn't changed,” said Daniel Clemens, owner of Packetninjas, in a recent Dark Reading report on low-level memory threats. "We still have code, we still have data. Exploiting memory corruption vulnerabilities is effectively flipping data to code for creative execution."

So far, however, there is little industry data to back up experts' suspicions about in-memory threats because most security analysis tools focus on stored data. Triumfant hopes its new research will help identify in-memory attacks and provide trend data over time.

"So far, we've only tested it in our own environment, but we've been able to see a clear pattern," Prisco says. "System calls that take 20 or 25 milliseconds consistently go up to 50 milliseconds or more when there's an in-memory attack. When you have processing delays like that -- delays that are two or three deltas beyond the norm -- then you know that something is not right."

Triumfant is also working on a way to identify the memory objects responsible for the delays and remove them before they can execute, Prisco says.

"These in-memory attacks are going to become more attractive to the bad guys as conventional malware detection tools get better," Prisco predicts. "It's a way to execute the same attacks without being detected."

Have a comment on this story? Please click "Add a Comment" below. If you'd like to contact Dark Reading's editors directly, send us a message.

About the Author(s)

Tim Wilson, Editor in Chief, Dark Reading


Tim Wilson is Editor in Chief and co-founder of Dark, UBM Tech's online community for information security professionals. He is responsible for managing the site, assigning and editing content, and writing breaking news stories. Wilson has been recognized as one of the top cyber security journalists in the US in voting among his peers, conducted by the SANS Institute. In 2011 he was named one of the 50 Most Powerful Voices in Security by SYS-CON Media.

Keep up with the latest cybersecurity threats, newly discovered vulnerabilities, data breach information, and emerging trends. Delivered daily or weekly right to your email inbox.

You May Also Like

More Insights