Dark Reading is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them.Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Cloud

3/21/2018
11:00 AM
Connect Directly
Twitter
RSS
E-Mail
50%
50%

Online Sandboxing: A Stash for Exfiltrated Data?

SafeBreach researchers extend leaky sandbox research to show how services like VirusTotal and Hybrid Analysis could be used to steal data from air-gapped systems.

Online sandboxing services are an extremely important resource for threat hunters to use to consolidate knowledge and collaborate about emerging threats. However, new research out this week shows that the tables can be turned on organizations using services like VirusTotal or Hybrid Analysis by using them as a way to exfiltrate data from heavily locked down systems. 

Published by researchers with SafeBreach, the report released today adds a new layer to previous work the team did on leaky sandbox services used by cloud-based antivirus platforms. Presented at BlackHat last summer, that earlier research demonstrated how researchers could exfiltrate data off of air-gapped and otherwise disconnected systems by tricking cloud-enhanced anti virus agents into sending stolen data onto a criminal's server. 

This time around the SafeBreach team used that same principle of hiding a wolf in sheep's clothing to offer a proof-of-concept for how online sandboxing services could be used as an outbound channel for stolen sensitive data. Unlike the previous research, this time around the hidden data never needs for an attacker to communicate out of the sandbox in order for them to find their stash.

Here is the way it works: a piece of custom malware is designed to be found on the system and uploaded to an online sandboxing service like VirusTotal, either by a human security analyst or a tool integrated with the online service. Unbeknownst to the victim, the stolen data is encoded and embedded in that malware. And so is a secret and searchable passphrase.

"One of the main objectives that needs to be addressed by the attacker is how to make the data that they incorporate into the malware visible to him - to be identified by him after the malware is uploaded to the server - but still be hidden enough that any other user cannot identify that the data is (taken) from an organization, or from any malicious contents inside it," says Dor Azouri, one of the primary researchers for SafeBreach on this project.

Azouri explains that this can be achieved by developing a "magic" string of text and either encrypting or encoding it in such a way that the attacker can make a simple query into the public sandboxing platform in order to pull up that specific piece of malware that's been uploaded to the database.

By passively collecting the data in this way, the attacker doesn't need to emit outbound network traffic or run an HTTP server or authoritative DNS server, according to Azouri, noting that the process is very low-profile and can be done on even extremely isolated systems. The downside, he says, is that the attacker must know - or at least bet heavily - that their targeted victim is using an online sandbox engine as part of their normal security processes.

According to a VirusTotal spokesperson, the SafeBreach research presented here isn't a bug in VirusTotal. It is just a way for attackers to misuse the same kind of feature you'd have in any number of services widely used by enterprises today that allow the uploading, publishing and retrieving of data. Of course, the point from SafeBreach is that this method is meant to be used as a last-mile connection between an attacker and an otherwise well-fortified system. A sensitive system like this wouldn't be using any other kind of SaaS-based information sharing platform, but the security team might not think twice about sending off a sample to VirusTotal from that system.

Nevertheless, VirusTotal's spokesperson contends that this isn't a likely or practical channel for attackers.

"The approach laid out by Safebreach does not make practical sense from an attacker perspective," the spokesperson says. "As an attacker, you don't want to use a channel that is being inspected by the whole security industry to exfiltrate data when there are thousands of other public internet services that could accomplish the same thing much more discreetly.”

Related Content:

 Interop ITX 2018

Join Dark Reading LIVE for two cybersecurity summits at Interop ITX. Learn from the industry’s most knowledgeable IT security experts. Check out the security track here. Early Bird Rates expire Friday March 23. Use Promo Code DR200 & save $200.

Ericka Chickowski specializes in coverage of information technology and business innovation. She has focused on information security for the better part of a decade and regularly writes about the security industry as a contributor to Dark Reading.  View Full Bio
 

Recommended Reading:

Comment  | 
Print  | 
More Insights
Comments
Threaded  |  Newest First  |  Oldest First
COVID-19: Latest Security News & Commentary
Dark Reading Staff 9/25/2020
Hacking Yourself: Marie Moe and Pacemaker Security
Gary McGraw Ph.D., Co-founder Berryville Institute of Machine Learning,  9/21/2020
Startup Aims to Map and Track All the IT and Security Things
Kelly Jackson Higgins, Executive Editor at Dark Reading,  9/22/2020
Register for Dark Reading Newsletters
White Papers
Video
Cartoon
Current Issue
Special Report: Computing's New Normal
This special report examines how IT security organizations have adapted to the "new normal" of computing and what the long-term effects will be. Read it and get a unique set of perspectives on issues ranging from new threats & vulnerabilities as a result of remote working to how enterprise security strategy will be affected long term.
Flash Poll
How IT Security Organizations are Attacking the Cybersecurity Problem
How IT Security Organizations are Attacking the Cybersecurity Problem
The COVID-19 pandemic turned the world -- and enterprise computing -- on end. Here's a look at how cybersecurity teams are retrenching their defense strategies, rebuilding their teams, and selecting new technologies to stop the oncoming rise of online attacks.
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2020-15208
PUBLISHED: 2020-09-25
In tensorflow-lite before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, when determining the common dimension size of two tensors, TFLite uses a `DCHECK` which is no-op outside of debug compilation modes. Since the function always returns the dimension of the first tensor, malicious attackers can ...
CVE-2020-15209
PUBLISHED: 2020-09-25
In tensorflow-lite before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, a crafted TFLite model can force a node to have as input a tensor backed by a `nullptr` buffer. This can be achieved by changing a buffer index in the flatbuffer serialization to convert a read-only tensor to a read-write one....
CVE-2020-15210
PUBLISHED: 2020-09-25
In tensorflow-lite before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, if a TFLite saved model uses the same tensor as both input and output of an operator, then, depending on the operator, we can observe a segmentation fault or just memory corruption. We have patched the issue in d58c96946b and ...
CVE-2020-15211
PUBLISHED: 2020-09-25
In TensorFlow Lite before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, saved models in the flatbuffer format use a double indexing scheme: a model has a set of subgraphs, each subgraph has a set of operators and each operator has a set of input/output tensors. The flatbuffer format uses indices f...
CVE-2020-15212
PUBLISHED: 2020-09-25
In TensorFlow Lite before versions 2.2.1 and 2.3.1, models using segment sum can trigger writes outside of bounds of heap allocated buffers by inserting negative elements in the segment ids tensor. Users having access to `segment_ids_data` can alter `output_index` and then write to outside of `outpu...