Archive

Author Archive

How artificial intelligence stopped an Emotet outbreak

February 14th, 2018 No comments

At 12:46 a.m. local time on February 3, a Windows 7 Pro customer in North Carolina became the first would-be victim of a new malware attack campaign for Trojan:Win32/Emotet. In the next 30 minutes, the campaign tried to attack over a thousand potential victims, all of whom were instantly and automatically protected by Windows Defender AV.

How did Windows Defender AV uncover the newly launched attack and block it at the outset? Through layered machine learning, including use of both client-side and cloud machine learning (ML) models. Every day, artificial intelligence enables Windows Defender AV to stop countless malware outbreaks in their tracks. In this blog post, well take a detailed look at how the combination of client and cloud ML models detects new outbreaks.

Figure 1. Layered detected model in Windows Defender AV

Client machine learning models

The first layer of machine learning protection is an array of lightweight ML models built right into the Windows Defender AV client that runs locally on your computer. Many of these models are specialized for file types commonly abused by malware authors, including, JavaScript, Visual Basic Script, and Office macro. Some models target behavior detection, while other models are aimed at detecting portable executable (PE) files (.exe and .dll).

In the case of the Emotet outbreak on February 3, Windows Defender AV caught the attack using one of the PE gradient boosted tree ensemble models. This model classifies files based on a featurization of the assembly opcode sequence as the file is emulated, allowing the model to look at the files behavior as it was simulated to run.

Figure 2. A client ML model classified the Emotet outbreak as malicious based on emulated execution opcode machine learning model.

The tree ensemble was trained using LightGBM, a Microsoft open-source framework used for high-performance gradient boosting.

Figure 3a. Visualization of the LightBGM-trained client ML model that successfully classified Emotet’s emulation behavior as malicious. A set of 20 decision trees are combined in this model to classify whether the files emulated behavior sequence is malicious or not.

Figure 3b. A more detailed look at the first decision tree in the model. Each decision is based on the value of a different feature. Green triangles indicate weighted-clean decision result; red triangles indicate weighted malware decision result for the tree.

When the client-based machine learning model predicts a high probability of maliciousness, a rich set of feature vectors is then prepared to describe the content. These feature vectors include:

  • Behavior during emulation, such as API calls and executed code
  • Similarity fuzzy hashes
  • Vectors of content descriptive flags optimized for use in ML models
  • Researcher-driven attributes, such as packer technology used for obfuscation
  • File name
  • File size
  • Entropy level
  • File attributes, such as number of sections
  • Partial file hashes of the static and emulated content

This set of features form a signal sent to the Windows Defender AV cloud protection service, which runs a wide array of more complex models in real-time to instantly classify the signal as malicious or benign.

Real-time cloud machine learning models

Windows Defender AVs cloud-based real-time classifiers are powerful and complex ML models that use a lot of memory, disk space, and computational resources. They also incorporate global file information and Microsoft reputation as part of the Microsoft Intelligent Security Graph (ISG) to classify a signal. Relying on the cloud for these complex models has several benefits. First, it doesnt use your own computers precious resources. Second, the cloud allows us to take into consideration the global information and reputation information from ISG to make a better decision. Third, cloud-based models are harder for cybercriminals to evade. Attackers can take a local client and test our models without our knowledge all day long. To test our cloud-based defenses, however, attackers have to talk to our cloud service, which will allow us to react to them.

The cloud protection service is queried by Windows Defender AV clients billions of times every day to classify signals, resulting in millions of malware blocks per day, and translating to protection for hundreds of millions of customers. Today, the Windows Defender AV cloud protection service has around 30 powerful models that run in parallel. Some of these models incorporate millions of features each; most are updated daily to adapt to the quickly changing threat landscape. All together, these classifiers provide an array of classifications that provide valuable information about the content being scanned on your computer.

Classifications from cloud ML models are combined with ensemble ML classifiers, reputation-based rules, allow-list rules, and data in ISG to come up with a final decision on the signal. The cloud protection service then replies to the Windows Defender client with a decision on whether the signal is malicious or not all in a fraction of a second.

Figure 4. Windows Defender AV cloud protection service workflow.

In the Emotet outbreak, one of our cloud ML servers in North America received the most queries from customers; corresponding to where the outbreak began. At least nine real-time cloud-based ML classifiers correctly identified the file as malware. The cloud protection service replied to signals instructing the Windows Defender AV client to block the attack using two of our ML-based threat names, Trojan:Win32/Fuerboos.C!cl and Trojan:Win32/Fuery.A!cl.

This automated process protected customers from the Emotet outbreak in real-time. But Windows Defender AVs artificial intelligence didnt stop there.

Deep learning on the full file content

Automatic sample submission, a Windows Defender AV feature, sent a copy of the malware file to our backend systems less than a minute after the very first encounter. Deep learning ML models immediately analyzed the file based on the full file content and behavior observed during detonation. Not surprisingly, deep neural network models identified the file as a variant of Trojan:Win32/Emotet, a family of banking Trojans.

While the ML classifiers ensured that the malware was blocked at first sight, deep learning models helped associate the threat with the correct malware family. Customers who were protected from the attack can use this information to understand the impact the malware might have had if it were not stopped.

Additionally, deep learning models provide another layer of protection: in relatively rare cases where real-time classifiers are not able to come to a conclusive decision about a file, deep learning models can do so within minutes. For example, during the Bad Rabbit ransomware outbreak, Windows Defender AV protected customers from the new ransomware just 14 minutes after the very first encounter.

Intelligent real-time protection against modern threats

Machine learning and AI are at the forefront of the next-gen real-time protection delivered by Windows Defender AV. These technologies, backed by unparalleled optics into the threat landscape provided by ISG as well as world-class Windows Defender experts and researchers, allow Microsoft security products to quickly evolve and scale to defend against the full range of attack scenarios.

Cloud-delivered protection is enabled in Windows Defender AV by default. To check that its running, go to Windows Settings > Update & Security > Windows Defender. Click Open Windows Defender Security Center, then navigate to Virus & threat protection > Virus &threat protection settings, and make sure that Cloud-delivered protection and Automatic sample submission are both turned On.

In enterprise environments, the Windows Defender AV cloud protection service can be managed using Group Policy, System Center Configuration Manager, PowerShell cmdlets, Windows Management Instruction (WMI), Microsoft Intune, or via the Windows Defender Security Center app.

The intelligent real-time defense in Windows Defender AV is part of the next-gen security technologies in Windows 10 that protect against a wide spectrum of threats. Of particular note, Windows 10 S is not affected by this type of malware attack. Threats like Emotet wont run on Windows 10 S because it exclusively runs apps from the Microsoft Store. Learn more about Windows 10 S. To know about all the security technologies available in Windows 10, read Microsoft 365 security and management features available in Windows 10 Fall Creators Update.

 

Geoff McDonald, Windows Defender Research
with Randy Treit and Allan Sepillo

 

 


Talk to us

Questions, concerns, or insights on this story? Join discussions at the Microsoft community and Windows Defender Security Intelligence.

Follow us on Twitter @WDSecurity and Facebook Windows Defender Security Intelligence.

Protecting customers from being intimidated into making an unnecessary purchase

January 30th, 2018 No comments

There has been an increase in free versions of programs that purport to scan computers for various errors, and then use alarming, coercive messages to scare customers into buying a premium version of the same program. The paid version of these programs, usually called cleaner or optimizer applications, purportedly fixes the problems discovered by the free version. We find this practice problematic because it can pressure customers into making unnecessary purchase decisions.

To help protect customers from receiving such coercive messaging, we are updating our evaluation criteria to specify that programs must not use alarming or coercive messaging that can put pressure on customers into making a purchase or performing other actions. We use the evaluation criteria to determine what programs are identified as malware and unwanted software. In the future, programs that display coercive messaging will be classified as unwanted software, detected, and removed.

This update comes in addition to our other long-standing customer protection requirements designed to keep our customers from being deceived by programs that display misleading, exaggerated, or threatening messages about a systems health. In February 2016, we required cleaner and optimizer programs that purport to clean up systems and optimize performance to provide customers with detailed information about what purportedly needs to be fixed. This requirement aims to protect customers from programs that present aggregate “error results with no specific details, without providing customers with the ability to assess and validate the so-called errors.

We have recently updated our evaluation criteria to state:

Unwanted behaviors: coercive messaging

Programs must not display alarming or coercive messages or misleading content to pressure you into paying for additional services or performing superfluous actions.

Software that coerces users may display the following characteristics, among others:

  • Reports errors in an exaggerated or alarming manner about the users system and requires the user to pay for fixing the errors or issues monetarily or by performing other actions such as taking a survey, downloading a file, signing up for a newsletter, etc.
  • Suggests that no other actions will correct the reported errors or issues
  • Requires the user to act within a limited period of time to get the purported issue resolved

Starting March 1, 2018, Windows Defender Antivirus and other Microsoft security products will classify programs that display coercive messages as unwanted software, which will be detected and removed. If you are software developer and want to validate the detection of your programs, visit the Windows Defender Security Intelligence portal.

Customer protection is our top priority. We adjust, expand, and update our evaluation criteria based on customer feedback and in order to capture the latest developments in unwanted software and other threats. We encourage our customers to submit programs that exhibit unwanted behaviors related to coercive messaging, or other unwanted or malicious behaviors in general.

 

Barak Shein
Windows Defender Security Research

 


Talk to us

Questions, concerns, or insights on this story? Join discussions at the Microsoft community and Windows Defender Security Intelligence.

Follow us on Twitter @WDSecurity and Facebook Windows Defender Security Intelligence.

Categories: cybersecurity Tags:

Now you see me: Exposing fileless malware

January 24th, 2018 No comments

Attackers are determined to circumvent security defenses using increasingly sophisticated techniques. Fileless malware boosts the stealth and effectiveness of an attack, and two of last years major ransomware outbreaks (Petya and WannaCry) used fileless techniques as part of their kill chains.

The idea behind fileless malware is simple: If tools already exist on a device (for example PowerShell.exe or wmic.exe) to fulfill an attackers objectives, then why drop custom tools that could be flagged as malware? If an attacker can take over a process, run code in its memory space, and then use that code to call tools that are already on a device, the attack becomes more difficult to detect.

Successfully using this approach, sometimes called living off the land, is not a walk in the park. Theres another thing that attackers need to deal with: Establishing persistence. Memory is volatile, and with no files on disk, how can attackers get their code to auto-start after a system reboot and retain control of a compromised system?

Misfox: A fileless gateway to victim networks

In April 2016, a customer contacted the Microsoft Incident Response team about a case of cyber-extortion. The attackers had requested a substantial sum of money from the customer in exchange for not releasing their confidential corporate information that the attackers had stolen from the customers compromised computers. In addition, the attackers had threatened to “flatten” the network if the customer contacted law enforcement. It was a difficult situation.

Quick fact
Windows Defender AV detections of Misfox more than doubled in Q2 2017 compared to Q1 2017.

The Microsoft Incident Response team investigated machines in the network, identified targeted implants, and mapped out the extent of the compromise. The customer was using a well-known third-party antivirus product that was installed on the vast majority of machines. While it was up-to-date with the latest signatures, the AV product had not detected any targeted implants.

The Microsoft team then discovered that the attackers attempted to encrypt files with ransomware twice. Luckily, those attempts failed. As it turned out, the threat to flatten the network was a plan B to monetize the attack after their plan A had failed.

Whats more, the team also discovered that the attackers had covertly persisted in the network for at least seven months through two separate channels:

  • The first channel involved a backdoor named Swrort.A that was deployed on several machines; this backdoor was easily detected by antivirus.
  • The second channel was much more subtle and interesting, because:

    • It did not infect any files on the device
    • It left no artifacts on disk
    • Common file scanning techniques could not detect it

Should you disable PowerShell?
No. PowerShell is a powerful and secure management tool and is important for many system and IT functions. Attackers use malicious PowerShell scripts as post-exploitation technique that can only take place after an initial compromise has already occurred. Its misuse is a symptom of an attack that begins with other malicious actions like software exploitation, social engineering, or credential theft. The key is to prevent an attacker from getting into the position where they can misuse PowerShell. For tips on mitigating PowerShell abuse, continue reading.

The second tool was a strain of fileless malware called Misfox. Once Misfox was running in memory, it:

  • Created a registry run key that launches a “one-liner” PowerShell cmdlet
  • Launched an obfuscated PowerShell script stored in the registry BLOB; the obfuscated PowerShell script contained a reflective portable executable (PE) loader that loaded a Base64-encoded PE from the registry

Misfox did not drop any executable files, but the script stored in the registry ensured the malware persisted.

Fileless techniques

Misfox exemplifies how cyberattacks can incorporate fileless components in the kill chain. Attackers use several fileless techniques that can make malware implants stealthy and evasive. These techniques include:

  1. Reflective DLL injection
    Reflective DLL injection involves the manual loading of malicious DLLs into a process’ memory without the need for said DLLs to be on disk. The malicious DLL can be hosted on a remote attacker-controlled machine and delivered through a staged network channel (for example, Transport Layer Security (TLS) protocol), or embedded in obfuscated form inside infection vectors like macros and scripts. This results in the evasion of the OS mechanism that monitors and keeps track of loading executable modules. An example of malware that uses Reflective DLL injection is HackTool:Win32/Mikatz!dha.
  2. Memory exploits
    Adversaries use fileless memory exploits to run arbitrary code remotely on victim machines. For example, the UIWIX threat uses the EternalBlue exploit, which was used by both Petya and WannaCry, and has been observed to install the DoublePulsar backdoor, which lives entirely in the kernel’s memory (SMB Dispatch Table). Unlike Petya and Wannacry, UIWIX does not drop any files on disk.
  3. Script-based techniques
    Scripting languages provide powerful means for delivering memory-only executable payloads. Script files can embed encoded shellcodes or binaries that they can decrypt on the fly at run time and execute via .NET objects or directly with APIs without requiring them to be written to disk. The scripts themselves can be hidden in the registry (as in the case of Misfox), read from network streams, or simply run manually in the command-line by an attacker, without ever touching the disk.
  4. WMI persistence
    Weve seen certain attackers use the Windows Management Instrumentation (WMI) repository to store malicious scripts that are then invoked periodically using WMI bindings. This article [PDF] presents very good examples.

Fileless malware-specific mitigations on Microsoft 365

Microsoft 365 brings together a set of next-gen security technologies to protect devices, SaaS apps, email, and infrastructure from a wide spectrum of attacks. The following Windows-related components from Microsoft 365 have capabilities to detect and mitigate malware that rely on fileless techniques:

Tip
In addition to fileless malware-specific mitigations, Windows 10 comes with other next-gen security technologies that mitigate attacks in general. For example, Windows Defender Application Guard can stop the delivery of malware, fileless or otherwise, through Microsoft Edge and Internet Explorer. Read about the Microsoft 365 security and management features available in Windows 10 Fall Creators Update.

Windows Defender Antivirus

Windows Defender AV blocks the vast majority of malware using generic, heuristic, and behavior-based detections, as well as local and cloud-based machine learning models. Windows Defender AV protects against fileless malware through these capabilities:

  • Detecting script-based techniques by leveraging AMSI, which provides the capability to inspect PowerShell and other script types, even with multiple layers of obfuscation
  • Detecting and remediating WMI persistence techniques by scanning the WMI repository, both periodically and whenever anomalous behavior is observed
  • Detecting reflective DLL injection through enhanced memory scanning techniques and behavioral monitoring

Windows Defender Exploit Guard

Windows Defender Exploit Guard (Windows Defender EG), a new set of host intrusion prevention capabilities, helps reduce the attack surface area by locking down the device against a wide variety of attack vectors. It can help stop attacks that use fileless malware by:

  • Mitigating kernel-memory exploits like EternalBlue through Hypervisor Code Integrity (HVCI), which makes it extremely difficult to inject malicious code using kernel-mode software vulnerabilities
  • Mitigating user-mode memory exploits through the Exploit protection module, which consists of a number of exploit mitigations that can be applied either at the operating system level or at the individual app level
  • Mitigating many script-based fileless techniques, among other techniques, through Attack Surface Reduction (ASR) rules that lock down application behavior

Tip
On top of technical controls, it is important that administrative controls related to people and processes are also in place. The use of fileless techniques that rely on PowerShell and WMI on a remote victim machine requires that the adversary has privileged access to those machines. This may be due to poor administrative practices (for example, configuring a Windows service to run in the context of a domain admin account) that can enable credential theft. Read more about Securing Privileged Access.

Windows Defender Application Control

Windows Defender Application Control (WDAC) offers a mechanism to enforce strong code Integrity policies and to allow only trusted applications to run. In the context of fileless malware, WDAC locks down PowerShell to Constrained Language Mode, which limits the extended language features that can lead to unverifiable code execution, such as direct .NET scripting, invocation of Win32 APIs via the Add-Type cmdlet, and interaction with COM objects. This essentially mitigates PowerShell-based reflective DLL injection attacks.

Windows Defender Advanced Threat Protection

Windows Defender Advanced Threat Protection (Windows Defender ATP) is the integrated platform for our Windows Endpoint Protection (EPP) and Endpoint Detection and Response (EDR) capabilities. When it comes to post breach scenarios ATP alerts enterprise customers about highly sophisticated and advanced attacks on devices and corporate networks that other preventive protection features have been unable to defend against. It uses rich security data, advanced behavioral analytics, and machine learning to detect such attacks. It can help detect fileless malware in a number of ways, including:

  • Exposing covert attacks that use fileless techniques like reflective DLL loading using specific instrumentations that detect abnormal memory allocations
  • Detecting script-based fileless attacks by leveraging AMSI, which provides runtime inspection capability into PowerShell and other script-based malware, and applying machine learning models

Microsoft Edge

According to independent security tester NSS Labs, Microsoft Edge blocks more phishing sites and socially engineered malware than other browsers. Microsoft Edge mitigates fileless malware using arbitrary code protection capabilities, which can prevent arbitrary code, including malicious DLLs, from running. This helps mitigate reflective DLL loading attacks. In addition, Microsoft Edge offers a wide array of protections that mitigate threats, fileless or otherwise, using Windows Defender Application Guard integration and Windows Defender SmartScreen.

Windows 10 S

Windows 10 S is a special configuration of Windows 10 that combines many of the security features of Microsoft 365 automatically configured out of the box. It reduces attack surface by only allowing apps from the Microsoft Store. In the context of fileless malware, Windows 10 S has PowerShell Constrained Language Mode enabled by default. In addition, industry-best Microsoft Edge is the default browser, and Hypervisor Code Integrity (HVCI) is enabled by default.

 

Zaid Arafeh

Senior Program Manager, Windows Defender Research team

 


Talk to us

Questions, concerns, or insights on this story? Join discussions at the Microsoft community and Windows Defender Security Intelligence.

Follow us on Twitter @WDSecurity and Facebook Windows Defender Security Intelligence.

Detonating a bad rabbit: Windows Defender Antivirus and layered machine learning defenses

December 11th, 2017 No comments

Windows Defender Antivirus uses a layered approach to protection: tiers of advanced automation and machine learning models evaluate files in order to reach a verdict on suspected malware. While Windows Defender AV detects a vast majority of new malware files at first sight, we always strive to further close the gap between malware release and detection.

In a previous blog post, we looked at a real-world case study showing how Windows Defender Antivirus cloud protection service leverages next-gen security technologies to save “patient zero” from new malware threats in real-time. In that case study, a new Spora ransomware variant was analyzed and blocked within seconds using a deep neural network (DNN) machine learning classifier in the cloud. In this blog post well look at how additional automated analysis and machine learning models can further protect customers within minutes in rare cases where initial classification is inconclusive.

Layered machine learning models

In Windows Defender AVs layered approach to defense, if the first layer doesnt detect a threat, we move on to the next level of inspection. As we move down the layers, the amount of time required increases. However, we catch the vast majority of malware at the first (fastest) protection layers and only need to move on to a more sophisticated (but slower) level of inspection for rarer/more advanced threats.

For example, the vast majority of scanned objects are evaluated by the local Windows Defender client machine learning models, behavior-based detection algorithms, generic and heuristic classifications, and more. This helps ensure that users get the best possible performance. In rare cases where local intelligence cant reach a definitive verdict, Windows Defender AV will use the cloud for deeper analysis.

Figure 1. Layered detection model

For a more detailed look at our approach to protection, see The evolution of malware prevention.

Detonation-based machine learning classification

We use a variety of machine learning models that use different algorithms to predict whether a certain file is malware. Some of these algorithms are binary classifiers that give a strict clean-or-malware verdict (0 or 1), while others are multi-class classifiers that provide a probability for each classification (malware, clean, potentially unwanted application, etc). Each machine learning model is trained against a set of different features (often thousands, sometimes hundreds of thousands) to learn to distinguish between different kinds of programs.

For the fastest classifiers in our layered stack, the features may include static attributes of the file combined with events (for example, API calls or behaviors) seen while the scanning engine emulates the file using dynamic translation. If the results from these models are inconclusive, well take an even more in-depth look at what the malware does by actually executing it in a sandbox and observing its run-time behavior. This is known as dynamic analysis, or detonation, and happens automatically whenever we receive a new suspected malware sample.

The activities seen in the sandbox machine (for example, registry changes, file creation/deletion, process injection, network connections, and so forth) are recorded and provided as features to our ML models. These models can then combine both the static features obtained from scanning the file with the dynamic features observed during detonation to arrive at an even stronger prediction.

Figure 2. Detonation-based machine learning classification

Ransom:Win32/Tibbar.A Protection in 14 minutes

On October 24, 2017, in the wake of recent ransomware outbreaks such as Wannacry and NotPetya, news broke of a new threat spreading, primarily in Ukraine and Russia: Ransom:Win32/Tibbar.A (popularly known as Bad Rabbit).

This threat is a good example of how detonation-based machine learning came into play to protect Windows Defender AV customers. First though, lets look at what happened to patient zero.

At 11:17 a.m. local time on October 24, a user running Windows Defender AV in St. Petersburg, Russia was tricked into downloading a file named FlashUtil.exe from a malicious website. Instead of a Flash update, the program was really the just-released Tibbar ransomware.

Windows Defender AV scanned the file and determined that it was suspicious. A query was sent to the cloud protection service, where several metadata-based machine learning models found the file suspicious, but not with a high enough probability to block. The cloud protection service requested that Windows Defender AV client to lock the file, upload it for processing, and wait for a decision.

Within a few seconds the file was processed, and sample-analysis-based ML models returned their conclusions. In this case, a multi-class deep neural network (DNN) machine learning classifier correctly classified the Tibbar sample as malware, but with only an 81.6% probability score. In order to avoid false positives, cloud protection service is configured by default to require at least 90% probability to block the malware (these thresholds are continually evaluated and fine-tuned to find the right balance between blocking malware while avoiding the blocking of legitimate programs). In this case, the ransomware was allowed to run.

Figure 3. Ransom:Win32/Tibbar.A ransom note

Detonation chamber

In the meantime, while patient zero and eight other unfortunate victims (in Ukraine, Russia, Israel, and Bulgaria) contemplated whether to pay the ransom, the sample was detonated and details of the system changes made by the ransomware were recorded.

Figure 4. Sample detonation events used by the machine learning model

As soon as the detonation results were available, a multi-class deep neural network (DNN) classifier that used both static and dynamic features evaluated the results and classified the sample as malware with 90.7% confidence, high enough for the cloud to start blocking.

When a tenth Windows Defender AV customer in the Ukraine was tricked into downloading the ransomware at 11:31 a.m. local time, 14 minutes after the first encounter, cloud protection service used the detonation-based malware classification to immediately block the file and protect the customer.

At this point the cloud protection service had “learned” that this file was malware. It now only required metadata from the client with the hash of the file to issue blocking decisions and protect customers. As the attack gained momentum and began to spread, Windows Defender AV customers with cloud protection enabled were protected. Later, a more specific detection was released to identify the malware as Ransom:Win32/Tibbar.A.

Closing the gap

While we feel good about Windows Defender AV’s layered approach to protection, digging deeper and deeper with automation and machine learning in order to finally reach a verdict on suspected malware, we are continually seeking to close the gap even further between malware release and protection. The cases where we cannot block at first sight are increasingly rare, but there is so much to be done. As our machine learning models are continuously updated and retrained, we are able to make better predictions over time. Yet malware authors will not rest, and the ever-changing threat landscape requires continuous investment in new and better technologies to detect new threats, but also to effectively differentiate the good from the bad.

What about systems that do get infected while detonation and classification are underway? One area that we’re actively investing in is advanced remediation techniques that will let us reach back out to those systems in an organization that were vulnerable and, if possible, get them back to a healthy state.

If you are organization that is willing to accept a higher false positive risk in exchange for stronger protection, you can configure the cloud protection level to tell the Windows Defender AV cloud protection service to take a more aggressive stance towards suspicious files, such as blocking at lower machine learning probability thresholds. In the Tibbar example above, for example, a configuration like this could have protected patient zero using the initial 81% confidence score, and not wait for the higher confidence (detonation-based) result that came later. You can also configure the cloud extended timeout to give the cloud protection service more time to evaluate a first-seen threat.

As another layer of real-time protection against ransomware, enable Controlled folder access, which is one of the features of the new Windows Defender Exploit Guard. Controlled folder access protects files from tampering by locking folders so that ransomware and other unauthorized apps cant access them.

For enterprises, Windows Defender Exploit Guards other features (Attack Surface Reduction, Exploit protection, and Network protection) further protect networks from advanced attacks. Windows Defender Advanced Threat Protection can also alert security operations personnel about malware activities in the network so that personnel can promptly investigate and respond to attacks.

For users running Windows 10 S, malware like Tibbar simply wont run. Windows 10 S provides advanced levels of security by exclusively running apps from the Microsoft Store. Threats such as Tibbar are non-issues for Windows 10 S users. Learn more about Windows 10 S.

New machine learning and AI techniques, in combination with both static and dynamic analysis, gives Windows Defender AV the ability to block more and more malware threats at first sight and, if that fails, learn as quickly as possible that something is bad and start blocking it. Using a layered approach, with different ML models at each layer, gives us the ability to target a wide variety of threats quickly while maintaining low false positive rates. As we gather more data about a potential threat, we can provide predictions with higher and higher confidence and take action accordingly. It is an exciting time to be in the fray.

 

Randy Treit

Senior Security Researcher, Windows Defender Research

 

 


Talk to us

Questions, concerns, or insights on this story? Join discussions at the Microsoft community and Windows Defender Security Intelligence.

Follow us on Twitter @WDSecurity and Facebook Windows Defender Security Intelligence.

 

Microsoft teams up with law enforcement and other partners to disrupt Gamarue (Andromeda)

December 4th, 2017 No comments

Today, with help from Microsoft security researchers, law enforcement agencies around the globe, in cooperation with Microsoft Digital Crimes Unit (DCU), announced the disruption of Gamarue, a widely distributed malware that has been used in networks of infected computers collectively called the Andromeda botnet.

The disruption is the culmination of a journey that started in December 2015, when the Microsoft Windows Defender research team and DCU activated a Coordinated Malware Eradication (CME) campaign for Gamarue. In partnership with internet security firm ESET, we performed in-depth research into the Gamarue malware and its infrastructure.

Our analysis of more than 44,000 malware samples uncovered Gamarues sprawling infrastructure. We provided detailed information about that infrastructure to law enforcement agencies around the world, including:

  • 1,214 domains and IP addresses of the botnets command and control servers
  • 464 distinct botnets
  • More than 80 associated malware families

The coordinated global operation resulted in the takedown of the botnets servers, disrupting one of the largest malware operations in the world. Since 2011, Gamarue has been distributing a plethora of other threats, including:

A global malware operation

For the past six years, Gamarue has been a very active malware operation that, until the takedown, showed no signs of slowing down. Windows Defender telemetry in the last six months shows Gamarues global prevalence.

Figure 1. Gamarues global prevalence from May to November 2017

While the threat is global, the list of top 10 countries with Gamarue encounters is dominated by Asian countries.

Figure 2. Top 10 countries with the most Gamarue encounters from May to November 2017

In the last six months, Gamarue was detected or blocked on approximately 1,095,457 machines every month on average.

Figure 3. Machines, IPs, and unique file encounters for Gamarue from May to November 2017; data does not include LNK detections

The Gamarue bot

Gamarue is known in the underground cybercrime market as Andromeda bot. A bot is a program that allows an attacker to take control of an infected machine. Like many other bots, Gamarue is advertised as a crime kit that hackers can purchase.

The Gamarue crime kit includes the following components:

  • Bot-builder, which builds the malware binary that infects computers
  • Command-and-control application, which is a PHP-based dashboard application that allows hackers to manage and control the bots
  • Documentation on how to create a Gamarue botnet

A botnet is a network of infected machines that communicate with command-and-control (C&C) servers, which are computer servers used by the hacker to control infected machines.

The evolution of the Gamarue bot has been the subject of many thorough analyses by security researchers. At the time of takedown, there were five known active Gamarue versions: 2.06, 2.07, 2.08, 2.09, and 2.10. The latest and the most active is version 2.10.

Gamarue is modular, which means that its functionality can be extended by plugins that are either included in the crime kit or available for separate purchase. The Gamarue plugins include:

  • Keylogger ($150) Used for logging keystrokes and mouse activity in order to steal user names and passwords, financial information, etc
  • Rootkit (included in crime kit) Injects rootkit codes into all processes running on a victim computer to give Gamarue persistence
  • Socks4/5 (included in crime kit) Turns victim computer into a proxy server for serving malware or malicious instructions to other computers on the internet
  • Formgrabber ($250) Captures any data submitted through web browsers (Chrome, Firefox, and Internet Explorer)
  • Teamviewer ($250) Enables attacker to remotely control the victim machine, spy on the desktop, perform file transfer, among other functions
  • Spreader Adds capability to spread Gamarue malware itself via removable drives (for example, portable hard drives or flash drives connected via a USB port); it also uses Domain Name Generation (DGA) for the servers where it downloads updates

Gamarue attack kill-chain

Over the years, various attack vectors have been used to distribute Gamarue. These include:

  • Removable drives
  • Social media (such as Facebook) messages with malicious links to websites that host Gamarue
  • Drive-by downloads/exploit kits
  • Spam emails with malicious links
  • Trojan downloaders

Once Gamarue has infected a machine, it contacts the C&C server, making the machine part of the botnet. Through the C&C server, the hacker can control Gamarue-infected machines, steal information, or issue commands to download additional malware modules.

Figure 4. Gamarues attack kill-chain

Gamarues main goal is to distribute other prevalent malware families. During the CME campaign, we saw at least 80 different malware families distributed by Gamarue. Some of these malware families include:

The installation of other malware broadens the scale of what hackers can do with the network of infected machines.

Command-and-control communication

When the Gamarue malware triggers the infected machine to contact the C&C server, it provides information like the hard disks volume serial number (used as the bot ID for the computer), the Gamarue build ID, the operating system of the infected machine, the local IP address, an indication whether the signed in user has administrative rights, and keyboard language setting for the infected machine. This information is sent to the C&C server via HTTP using the JSON format:

Figure 5. Information sent by Gamarue to C&C server

The information about keyboard language setting is very interesting, because the machine will not be further infected if the keyboard language corresponds to the following countries:

  • Belarus
  • Russia
  • Ukraine
  • Kazahkstan

Before sending to the C&C server, this information is encrypted with RC4 algorithm using a key hardcoded in the Gamarue malware body.

Figure 6. Encrypted C&C communication

Once the C&C server receives the message, it sends a command that is pre-assigned by the hacker in the control dashboard.

Figure 7. Sample control dashboard used by attackers to communicate to Gamarue bots

The command can be any of the following:

  • Download EXE (i.e., additional executable malware files)
  • Download DLL (i.e., additional malware; removed in version 2.09 and later)
  • Install plugin
  • Update bot (i.e., update the bot malware)
  • Delete DLLs (removed in version 2.09 and later)
  • Delete plugins
  • Kill bot

The last three commands can be used to remove evidence of Gamarue presence in machines.

The reply from the C&C server is also encrypted with RC4 algorithm using the same key used to encrypt the message from the infected machine.

Figure 8. Encrypted reply from C&C server

When decrypted, the reply contains the following information:

  • Time interval in minutes time to wait for when to ask the C2 server for the next command
  • Task ID – used by the hacker to track if there was an error performing the task
  • Command one of the command mentioned above
  • Download URL – from which a plugin/updated binary/other malware can be downloaded depending on the command.

Figure 9. Decrypted reply from C&C server

Anti-sandbox techniques

Gamarue employs anti-AV techniques to make analysis and detection difficult. Prior to infecting a machine, Gamarue checks a list hashes of the processes running on a potential victims machine. If it finds a process that may be associated with malware analysis tools, such as virtual machines or sandbox tools, Gamarue does not infect the machine. In older versions, a fake payload is manifested when running in a virtual machine.

Figure 10. Gamarue checks if any of the running processes are associated with malware analysis tools

Stealth mechanisms

Gamarue uses cross-process injection techniques to stay under the radar. It injects its code into the following legitimate processes:

  • msiexec.exe (Gamarue versions 2.07 to 2.10)
  • wuauclt.exe, wupgrade.exe, svchost.exe (version 2.06)

It can also use a rootkit plugin to hide the Gamarue file and its autostart registry entry.

Gamarue employs a stealthy technique to store and load its plugins as well. The plugins are stored fileless, either saved in the registry or in an alternate data stream of the Gamarue file.

OS tampering

Gamarue attempts to tamper with the operating systems of infected computers by disabling Firewall, Windows Update, and User Account Control functions. These functionalities cannot be re-enabled until the Gamarue infection has been removed from the infected machine. This OS tampering behavior does not work on Windows 10

Figure 11. Disabled Firewall and Windows Update

Monetization

There are several ways hackers earn using Gamarue. Since Gamarues main purpose is to distribute other malware, hackers earn using pay-per-install scheme. Using its plugins, Gamarue can also steal user information; stolen information can be sold to other hackers in cybercriminal underground markets. Access to Gamarue-infected machines can also be sold, rented, leased, or swapped by one criminal group to another.

Remediation

To help prevent a Gamarue infection, as well as other malware and unwanted software, take these precautions:

  • Be cautious when opening emails or social media messages from unknown users.
  • Be wary about downloading software from websites other than the program developers.

More importantly, ensure you have the right security solutions that can protect your machine from Gamarue and other threats. Windows Defender Antivirus detects and removes the Gamarue malware. With advanced machine learning models, as well as generic and heuristic techniques, Windows Defender AV detects new as well as never-before-seen malware in real-time via the cloud protection service. Alternatively, standalone tools, such as Microsoft Safety Scanner and the Malicious Software Removal Tool (MSRT), can also detect and remove Gamarue.

Microsoft Edge can block Gamarue infections from the web, such as those from malicious links in social media messages and drive-by downloads or exploit kits. Microsoft Edge is a secure browser that opens pages within low privilege app containers and uses reputation-based blocking of malicious downloads.

In enterprise environments, additional layers of protection are available. Windows Defender Advanced Threat Protection can help security operations personnel to detect Gamarue activities, including cross-process injection techniques, in the network so they can investigate and respond to attacks. Windows Defender ATPs enhanced behavioral and machine learning detection libraries flag malicious behavior across the malware infection process, from delivery and installation, to persistence mechanisms, and command-and-control communication.

Microsoft Exchange Online Protection (EOP) can block Gamarue infections from email uses built-in anti-spam filtering capabilities that help protect Office 365 customers. Office 365 Advanced Threat Protection helps secure mailboxes against email attacks by blocking emails with unsafe attachments, malicious links, and linked-to files leveraging time-of-click protection.

Windows Defender Exploit Guard can block malicious documents (such as those that distribute Gamarue) and scripts. The Attack Surface Reduction (ASR) feature in Windows Defender Exploit Guard uses a set of built-in intelligence that can block malicious behaviors observed in malicious documents. ASR rules can also be turned on to block malicious attachments from being run or launched from Microsoft Outlook or webmail (such as Gmail, Hotmail, or Yahoo).

Microsoft is also continuing the collaborative effort to help clean Gamarue-infected computers by providing a one-time package with samples (through the Virus Information Alliance) to help organizations protect their customers.

 

 

Microsoft Digital Crimes Unit and Windows Defender Research team

 

 

Get more info on the Gamarue (Andromeda) takedown from the following sources:

 

 


Talk to us

Questions, concerns, or insights on this story? Join discussions at the Microsoft community and Windows Defender Security Intelligence.

Follow us on Twitter @WDSecurity and Facebook Windows Defender Security Intelligence.

 

Windows Defender ATP machine learning and AMSI: Unearthing script-based attacks that ‘live off the land’

December 4th, 2017 No comments

Data center

Scripts are becoming the weapon of choice of sophisticated activity groups responsible for targeted attacks as well as malware authors who indiscriminately deploy commodity threats.

Scripting engines such as JavaScript, VBScript, and PowerShell offer tremendous benefits to attackers. They run through legitimate processes and are perfect tools for living off the landstaying away from the disk and using common tools to run code directly in memory. Often part of the operating system, scripting engines can evaluate and execute content from the internet on-the-fly. Furthermore, integration with popular apps make them effective vehicles for delivering malicious implants through social engineering as evidenced by the increasing use of scripts in spam campaigns.

Malicious scripts are not only used as delivery mechanisms. We see them in various stages of the kill chain, including during lateral movement and while establishing persistence. During these latter stages, the scripting engine of choice is clearly PowerShellthe de facto scripting standard for administrative tasks on Windowswith the ability to invoke system APIs and access a variety of system classes and objects.

While the availability of powerful scripting engines makes scripts convenient tools, the dynamic nature of scripts allows attackers to easily evade analysis and detection by antimalware and similar endpoint protection products. Scripts are easily obfuscated and can be loaded on-demand from a remote site or a key in the registry, posing detection challenges that are far from trivial.

Windows 10 provides optics into script behavior through Antimalware Scan Interface (AMSI), a generic, open interface that enables Windows Defender Antivirus to look at script contents the same way script interpreters doin a form that is both unencrypted and unobfuscated. In Windows 10 Fall Creators Update, with knowledge from years analyzing script-based malware, weve added deep behavioral instrumentation to the Windows script interpreter itself, enabling it to capture system interactions originating from scripts. AMSI makes this detailed interaction information available to registered AMSI providers, such as Windows Defender Antivirus, enabling these providers to perform further inspection and vetting of runtime script execution content.

This unparalleled visibility into script behavior is capitalized further through other Windows 10 Fall Creators Update enhancements in both Windows Defender Antivirus and Windows Defender Advanced Threat Protection (Windows Defender ATP). Both solutions make use of powerful machine learning algorithms that process the improved optics, with Windows Defender Antivirus delivering enhanced blocking of malicious scripts pre-breach and Windows Defender ATP providing effective behavior-based alerting for malicious post-breach script activity.

In this blog, we explore how Windows Defender ATP, in particular, makes use of AMSI inspection data to surface complex and evasive script-based attacks. We look at advanced attacks perpetrated by the highly skilled KRYPTON activity group and explore how commodity malware like Kovter abuses PowerShell to leave little to no trace of malicious activity on disk. From there, we look at how Windows Defender ATP machine learning systems make use of enhanced insight about script characteristics and behaviors to deliver vastly improved detection capabilities.

KRYPTON: Highlighting the resilience of script-based attacks

Traditional approaches for detecting potential breaches are quite file-centric. Incident responders often triage autostart entries, sorting out suspicious files by prevalence or unusual name-folder combinations. With modern attacks moving closer towards being completely fileless, it is crucial to have additional sensors at relevant choke points.

Apart from not having files on disk, modern script-based attacks often store encrypted malicious content separately from the decryption key. In addition, the final key often undergoes multiple processes before it is used to decode the actual payload, making it is impossible to make a determination based on a single file without tracking the actual invocation of the script. Even a perfect script emulator would fail this task.

For example, the activity group KRYPTON has been observed hijacking or creating scheduled tasksthey often target system tasks found in exclusion lists of popular forensic tools like Autoruns for Windows. KRYPTON stores the unique decryption key within the parameters of the scheduled task, leaving the actual payload content encrypted.

To illustrate KRYPTON attacks, we look at a tainted Microsoft Word document identified by John Lambert and the Office 365 Advanced Threat Protection team.

KRYPTON lure document

Figure 1. KRYPTON lure document

To live off the land, KRYPTON doesnt drop or carry over any traditional malicious binaries that typically trigger antimalware alerts. Instead, the lure document contains macros and uses the Windows Scripting Host (wscript.exe) to execute a JavaScript payload. This script payload executes only with the right RC4 decryption key, which is, as expected, stored as an argument in a scheduled task. Because it can only be triggered with the correct key introduced in the right order, the script payload is resilient against automated sandbox detonations and even manual inspection.

KRYPTON script execution chain through wscript.exe

Figure 2. KRYPTON script execution chain through wscript.exe

Exposing actual script behavior with AMSI

AMSI overcomes KRYPTONs evasion mechanisms by capturing JavaScript API calls after they have been decrypted and ready to be executed by the script interpreter. The screenshot below shows part of the exposed content from the KRYPTON attack as captured by AMSI.

Part of the KRYPTON script payload captured by AMSI and sent to the cloud for analysis

Figure 3. Part of the KRYPTON script payload captured by AMSI and sent to the cloud for analysis

By checking the captured script behavior against indicators of attack (IoAs) built up by human experts as well as machine learning algorithms, Windows Defender ATP effortlessly flags the KRYPTON scripts as malicious. At the same time, Windows Defender ATP provides meaningful contextual information, including how the script is triggered by a malicious Word document.

Windows Defender ATP machine learning detection of KRYPTON script captured by AMSI

Figure 4. Windows Defender ATP machine learning detection of KRYPTON script captured by AMSI

PowerShell use by Kovter and other commodity malware

Not only advanced activity groups like KRYPTON are shifting from binary executables to evasive scripts. In the commodity space, Kovter malware uses several processes to eventually execute its malicious payload. This payload resides in a PowerShell script decoded by a JavaScript (executed by wscript.exe) and passed to powershell.exe as an environment variable.

Windows Defender ATP machine learning alert for the execution of the Kovter script-based payload

Figure 5. Windows Defender ATP machine learning alert for the execution of the Kovter script-based payload

By looking at the PowerShell payload content captured by AMSI, experienced analysts can easily spot similarities to PowerSploit, a publicly available set of penetration testing modules. While such attack techniques involve file-based components, they remain extremely hard to detect using traditional methods because malicious activities occur only in memory. Such behavior, however, is effortlessly detected by Windows Defender ATP using machine learning that combines detailed AMSI signals with signals generated by PowerShell activity in general.

Part of the Kovter script payload captured by AMSI and sent to the cloud for analysis

Figure 6. Part of the Kovter script payload captured by AMSI and sent to the cloud for analysis

Fresh machine learning insight with AMSI

While AMSI provides rich information from captured script content, the highly variant nature of malicious scripts continues to make them challenging targets for detection. To efficiently extract and identify new traits differentiating malicious scripts from benign ones, Windows Defender ATP employs advanced machine learning methods.

As outlined in our previous blog, we employ a supervised machine learning classifier to identify breach activity. We build training sets based on malicious behaviors observed in the wild and normal activities on typical machines, augmenting that with data from controlled detonations of malicious artifacts. The diagram below conceptually shows how we capture malicious behaviors in the form of process trees.

Process tree augmented by instrumentation for AMSI data

Figure 7. Process tree augmented by instrumentation for AMSI data

As shown in the process tree, the kill chain begins with a malicious document that causes Microsoft Word (winword.exe) to launch PowerShell (powershell.exe). In turn, PowerShell executes a heavily obfuscated script that drops and executes the malware fhjUQ72.tmp, which then obtains persistence by adding a run key to the registry. From the process tree, our machine learning systems can extract a variety of features to build expert classifiers for areas like registry modification and file creation, which are then converted into numeric scores that are used to decide whether to raise alerts.

With the instrumentation of AMSI signals added as part of the Windows 10 Fall Creators Update (version 1709), Windows Defender ATP machine learning algorithms can now make use of insight into the unobfuscated script content while continually referencing machine state changes associated with process activity. Weve also built a variety of script-based models that inspect the nature of executed scripts, such as the count of obfuscation layers, entropy, obfuscation features, ngrams, and specific API invocations, to name a few.

As AMSI peels off the obfuscation layers, Windows Defender ATP benefits from growing visibility and insight into API calls, variable names, and patterns in the general structure of malicious scripts. And while AMSI data helps improve human expert knowledge and their ability to train learning systems, our deep neural networks automatically learn features that are often hidden from human analysts.

Machine-learning detections of JavaScript and PowerShell scripts

Figure 8. Machine learning detections of JavaScript and PowerShell scripts

While these new script-based machine learning models augment our expert classifiers, we also correlate new results with other behavioral information. For example, Windows Defender ATP correlates the detection of suspicious script contents from AMSI with other proximate behaviors, such as network connections. This contextual information is provided to SecOps personnel, helping them respond to incidents efficiently.

Machine learning combines VBScript content from AMSI and tracked network activity

Figure 9. Machine learning combines VBScript content from AMSI and tracked network activity

Detection of AMSI bypass attempts

With AMSI providing powerful insight into malicious script activity, attacks are more likely to incorporate AMSI bypass mechanisms that we group into three categories:

  • Bypasses that are part of the script content and can be inspected and alerted on
  • Tampering with the AMSI sensor infrastructure, which might involve the replacement of system files or manipulation of the load order of relevant DLLs
  • Patching of AMSI instrumentation in memory

The Windows Defender ATP research team proactively develops anti-tampering mechanisms for all our sensors. We have devised heuristic alerts for possible manipulation of our optics, designing these alerts so that they are triggered in the cloud before the bypass can suppress them.

During actual attacks involving CVE-2017-8759, Windows Defender ATP not only detected malicious post-exploitation scripting activity but also detected attempts to bypass AMSI using code similar to one identified by Matt Graeber.

Windows Defender ATP alert based on AMSI bypass pattern

Figure 10. Windows Defender ATP alert based on AMSI bypass pattern

AMSI itself captured the following bypass code for analysis in the Windows Defender ATP cloud.

AMSI bypass code sent to the cloud for analysis

Figure 11. AMSI bypass code sent to the cloud for analysis

Conclusion: Windows Defender ATP machine learning and AMSI provide revolutionary defense against highly evasive script-based attacks

Provided as an open interface on Windows 10, Antimalware Scan Interface delivers powerful optics into malicious activity hidden in encrypted and obfuscated scripts that are oftentimes never written to disk. Such evasive use of scripts is becoming commonplace and is being employed by both highly skilled activity groups and authors of commodity malware.

AMSI captures malicious script behavior by looking at script content as it is interpreted, without having to check physical files or being hindered by obfuscation, encryption, or polymorphism. At the endpoint, AMSI benefits local scanners, providing the necessary optics so that even obfuscated and encrypted scripts can be inspected for malicious content. Windows Defender Antivirus, specifically, utilizes AMSI to dynamically inspect and block scripts responsible for dropping all kinds of malicious payloads, including ransomware and banking trojans.

With Windows 10 Fall Creators Update (1709), newly added script runtime instrumentation provides unparalleled visibility into script behaviors despite obfuscation. Windows Defender Antivirus uses this treasure trove of behavioral information about malicious scripts to deliver pre-breach protection at runtime. To deliver post-breach defense, Windows Defender ATP uses advanced machine learning systems to draw deeper insight from this data.

Apart from looking at specific activities and patterns of activities, new machine learning algorithms in Windows Defender ATP look at script obfuscation layers, API invocation patterns, and other features that can be used to efficiently identify malicious scripts heuristically. Windows Defender ATP also correlates script-based indicators with other proximate activities, so it can deliver even richer contextual information about suspected breaches.

To benefit from the new script runtime instrumentation and other powerful security enhancements like Windows Defender Exploit Guard, customers are encourage to install Windows 10 Fall Creators Update.

Read the The Total Economic Impact of Microsoft Windows Defender Advanced Threat Protection from Forrester to understand the significant cost savings and business benefits enabled by Windows Defender ATP. To directly experience how Windows Defender ATP can help your enterprise detect, investigate, and respond to advance attacks, sign up for a free trial.

 

Stefan Sellmer, Windows Defender ATP Research

with

Shay Kels, Windows Defender ATP Research

Karthik Selvaraj, Windows Defender Research

 

Additional readings

 


Talk to us

Questions, concerns, or insights on this story? Join discussions at the Microsoft community and Windows Defender Security Intelligence.

Follow us on Twitter @WDSecurity and Facebook Windows Defender Security Intelligence.