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

11/11/2019
05:35 PM
Connect Directly
Twitter
LinkedIn
Google+
RSS
E-Mail
50%
50%

Researchers Find New Approach to Attacking Cloud Infrastructure

Cloud APIs' accessibility over the Internet opens a new window for adversaries to gain highly privileged access to cloud assets.

Public cloud infrastructure presents security teams with a new invisible management layer, creating new security challenges that demand better understanding. Many organizations don't properly understand the cloud identity and access management layer and often fail to secure it.

Such misunderstandings usually lead to dangerous misconfigurations that can drive customer risk; for example, in the case of the recent Capital One breach. Current security practices and controls are not sufficient to mitigate the risk posed by misunderstanding of the public cloud, explain Igal Gofman, XM head of security research, and Yaron Shani, XM senior security researcher.

When Gofman and Shani began to research cloud-focused threats, they realized many popular defense mechanisms focus on specific attack vectors: for example, brute force protections against cloud services and applications like password spray tools or AWS recon tools. Post-breach defense is usually based on different user activities and machine learning algorithms.

"The missing link in this approach is that those mechanisms are usually defensive in nature and usually not predictive," the researchers explain in an interview with Dark Reading. Traditional protections primarily focus on network, application, and operating system defense, they say.

A new attack vector exists in cloud providers' application programming interfaces (API), which are accessible through the Internet and give adversaries an opportunity to take advantage and gain highly privileged access to critical assets in the cloud. The people in charge of managing cloud resources are usually members of the DevOps, development, and IT teams, who gain access to APIs using different software development kits and dedicated command line tools.

"Once those account credentials are compromised, gaining access to high-value resources is trivial," the researchers say. Even if an organization makes a private subnet not open to the Internet, they add, cloud APIs can be easily accessed from the Internet with the right API key. Cloud provider tools—for example, the command-line interface tool (CLI) — save the user's credentials inside a file, which is typically locally stored on the individual's workstation.

At this year's Black Hat Europe, Gofman and Shani plan to demonstrate an alternative new approach to attacking cloud infrastructure in a talk titled "Inside Out — The Cloud Has Never Been So Close." Their methodology involves using a graph to show permission relationships between different entities, revealing risky choke points that need to be addressed and eliminated. The outcome of this graph, they say, can be used by red and blue teams to gain deeper understanding of permission relationships in cloud environments. After explaining the connections, they'll show how attackers can abuse features to gain privileges.

Attackers don't need to be sophisticated to take advantage of public cloud APIs, they say, noting they didn't find any open-source tools that automate the entire stack of the research.

"In practice, the sophistication required to develop such tools is not high, because basically all the information is publicly available and well-documented by most cloud providers, meaning they document each security feature in great detail and it can serve both the defenders and the adversaries," Gofman and Shani say. In general, they continue, developing an offensive tool that leverages their attack research would be easier than building a defensive system around it.

In terms of protecting themselves, the first and most important steps companies should follow are best practice guides from cloud providers, the researchers say. Large and complex organizations often have trouble tracking and monitoring permissions in large cloud infrastructures, and evaluating general organizational risk factors, they explain. They suggest constantly monitoring paths attacks can take to high-value cloud resources.  

Related Content:

Check out The Edge, Dark Reading's new section for features, threat data, and in-depth perspectives. Today's top story: "What a Security Products Blacklist Means for End Users and Integrators."

Kelly Sheridan is the Staff Editor at Dark Reading, where she focuses on cybersecurity news and analysis. She is a business technology journalist who previously reported for InformationWeek, where she covered Microsoft, and Insurance & Technology, where she covered financial ... View Full Bio
 

Recommended Reading:

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
greg.jensen
50%
50%
greg.jensen,
User Rank: Author
11/22/2019 | 12:40:49 PM
Moving beyond the API
One of the most understood areas of ITsec is the issues that a modern DevSecOps program is starting to expose. Many of today's most sophisticated business-critical applications (Finance, HR, supply chain, CRM) that are 100% cloud native, often are part of a hybrid model with back-end services that reside in the customer's data center.  This often creates the risk that a compromise of the cloud service in question sometimes has entitled access to non-cloud data repositories in the customer data center.  This is why a DevSecOps model must be designed to canvas the "hybrid cloud" with consistent policies and workflows. 
COVID-19: Latest Security News & Commentary
Dark Reading Staff 6/1/2020
Stay-at-Home Orders Coincide With Massive DNS Surge
Robert Lemos, Contributing Writer,  5/27/2020
Register for Dark Reading Newsletters
White Papers
Video
Cartoon Contest
Write a Caption, Win a Starbucks Card! Click Here
Latest Comment: "Well I dont run on MacOS, so I need to take extra precautions"
Current Issue
How Cybersecurity Incident Response Programs Work (and Why Some Don't)
This Tech Digest takes a look at the vital role cybersecurity incident response (IR) plays in managing cyber-risk within organizations. Download the Tech Digest today to find out how well-planned IR programs can detect intrusions, contain breaches, and help an organization restore normal operations.
Flash Poll
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2020-13659
PUBLISHED: 2020-06-02
address_space_map in exec.c in QEMU 4.2.0 can trigger a NULL pointer dereference related to BounceBuffer.
CVE-2020-10703
PUBLISHED: 2020-06-02
A NULL pointer dereference was found in the libvirt API responsible introduced in upstream version 3.10.0, and fixed in libvirt 6.0.0, for fetching a storage pool based on its target path. In more detail, this flaw affects storage pools created without a target path such as network-based pools like ...
CVE-2020-10739
PUBLISHED: 2020-06-02
Istio 1.4.x before 1.4.9 and Istio 1.5.x before 1.5.4 contain the following vulnerability when telemetry v2 is enabled: by sending a specially crafted packet, an attacker could trigger a Null Pointer Exception resulting in a Denial of Service. This could be sent to the ingress gateway or a sidecar, ...
CVE-2020-10136
PUBLISHED: 2020-06-02
Multiple products that implement the IP Encapsulation within IP standard (RFC 2003, STD 1) decapsulate and route IP-in-IP traffic without any validation, which could allow an unauthenticated remote attacker to route arbitrary traffic via an exposed network interface and lead to spoofing, access cont...
CVE-2020-13757
PUBLISHED: 2020-06-01
Python-RSA 4.0 ignores leading '\0' bytes during decryption of ciphertext. This could conceivably have a security-relevant impact, e.g., by helping an attacker to infer that an application uses Python-RSA, or if the length of accepted ciphertext affects application behavior (such as by causing exces...