My information Resource (blog.mir.net)

    Researchers at Palo Alto’s Unit 42 have discovered a worm that mines Monero, a privacy focused cryptocurrency, and spreads itself via infected Docker Daemons in the Docker Engine. Shodan scans of Docker engines show over 2000 unsecured Docker hosts. The researchers have named the cyptojacking malware Graboid. 

    Graboid has a downloader planted on an infected Docker image with a Docker Client tool used to connect to other Docker hosts. The attacker accesses an unsecured Docker host and infects it with the malicious image. Anti-virus solutions would normally look for viral content or virus like activity but not check the contents of data within container as the container is maintained separately from the main machine. This form of obfuscation has been observed in other containerization solutions before, but Graboid is exceptional in its erratic and relatively ineffective methodology.

    After retrieving and establishing the malicious image, the attacker then downloads the 4 shell scripts of DOOM. These Shell scrips are named live.sh, worm.sh, xmr.sh, and cleanxmr.sh. The first script, live.sh, surveys the victim assessing the resources to be plundered. it reports the number of available CPUs on the compromised host for the Command & control (C2) server to coordinate. The next script brings the ever hunting nose of the beast. The worm.sh script downloads the list of over 2000 vulnerable host’s IPs and replicates itself onto one of those IPs randomly. Then the last two scripts bring the chaos. The xmr.sh script deploys gakaws/nginx, a Monero cryptominer disguised to look like a NGIX load balancer/ web server, and does so on a randomly selected infected server. The last script, cleanxmr.sh, stops any xmrig based containers on another randomly selected infected server. It seems like Graboid runs Cleanxmr.sh before it runs xmr.sh as to avoid deactivating any Docker engines that just had their Monero mining capabilities turned on. This leads to a delay in the mining capabilities being turned on until the host is selected randomly by another infected host. Eventually the host will be selected to be disabled until a later time to be re enabled. This flash of infection and erratic appearances as well as the worm functionality has led to the researcher’s choice in naming the malware after the monsters in the 1990’s film Tremors.

    Graboid currently uses 15 C2 servers where 14 are included in the list of vulnerable IPs and the last has over 50 known vulnerabilities. The researchers have observed that it is likely these are controlled by the attacker illicitly. they have also calculated that it would have taken about 60 minutes to infect 70% of the vulnerable hosts with returns diminishing sharply after that. At that point there would be about 900 active miners at any particular time rotating through the available infected hosts with all of the infected hosts acting as nodes to facilitate communication with the Monero blockhain network. With a 100 second period of activity, a node is expected to be active for 250 seconds before being deactivated.

Sources:
 • https://unit42.paloaltonetworks.com/graboid-first-ever-cryptojacking-worm-found-in-images-on-docker-hub/

• https://threatpost.com/dockercontainers-graboid-cryptoworm/149235/

• https://securityaffairs.co/wordpress/92586/malware/graboidtargets-docker-hub.html

National Cyber Awareness System:

The National Security Agency (NSA) and the United Kingdom National Cyber
Security Centre (NCSC) have released a joint
advisory on advanced persistent threat (APT) group Turla—widely reported to
be Russian. The advisory provides an update to NCSC’s January 2018
report on Turla’s use of the malicious Neuron, Nautilus, and Snake tools to
steal sensitive data. Additionally, the advisory states that Turla has
compromised—and is currently leveraging—an Iranian APT group’s infrastructure
and resources, which include the Neuron and Nautilus tools.

The Cybersecurity and Infrastructure Security Agency (CISA) encourages users
and administrators to review the following resources for more information:
•    NSA Advisory Turla
Group Exploits Iranian APT To Expand Coverage Of Victims

•    UK NCSC Advisory Turla
group exploits Iranian APT to expand coverage of victims

•    January 2018 UK NCSC Report Turla Group Malware

    A commonly used method to secure network resources is a Virtual Private Network (VPN). They allow remote network devices to securely communicate with local resources as if they were physically plugged into the same network segment. You may even use one when working remotely to help keep your network traffic secure. While they can easily provide a lot of protection from various network attacks there are many pitfalls to avoid in order to keep the network resources secure.

    One common mistake when setting up a VPN is not properly securing the devices that the VPN provides access to. Because the servers or devices will not have direct inbound internet access many times a relaxed security policy is taken. This is because it is assumed that in order to access them an attacker would first have to either be on the network directly or be connected tuourhrough the VPN. Another common mistake is not regularly updating the VPN software. There are many reasons this can occur, including avoiding downtime or not wanting to break something that appears to be working fine as is.

    This week the National Security Agency (NSA) issued an advisory stating that APT groups have been actively using flaws in some popular VPN software to attack networks. They say the groups have weaponized three vulnerabilities against two pieces of VPN software, Pulse Secure VPN and Fortinet VPN. Two of the vulnerabilities, CVE-2019-11539 and CVE-201911510 specifically target Pulse Secure VPN servers. They allow remote unauthenticated command injection and arbitrary file reads on the VPN server device. The remaining vulnerability, CVE-201813379, targets Fortinet VPN servers and allows for remote unauthenticated arbitrary file reads from the server device. The National Cyber Security Center in the UK posted a separate advisory which added CVE-2018-13383 and CVE-2018-13383 to the list of vulnerabilities being used against Fortinet devices. Palo Alto Networks VPN software was also added to the vulnerable devices list with attackers utilizing CVE-2019-1579 for remote code execution on the affected VPN servers.

    In total the two agencies reported six vulnerabilities against three separate VPN software vendors. For each of the affected VPN products the vulnerabilities being used could allow an attacker access to the network resources as if the attacker were physically on the network. All of the affected products have updates available to fix these flaws so it is important that they are updated immediately if an affected version is still in use. The NSA also recommends rotating any existing VPN keys or tokens just in case they were stolen before the patches were able to be applied. 
Sources:

• https://threatpost.com/apt-groupsexploiting-flaws-in-unpatched-vpnsofficials-warn/148956/

• https://www.cyberscoop.com/vpnvulnerabilities-china-apt-palo-alto/

Original
release date: October 4, 2019

The Dutch National Cyber Security Centre (NCSC) has released a fact sheet on
the increasing difficulty of Domain Name System (DNS) monitoring. NCSC warns
that although modernization of transport protocols is helpful, it also makes it
more difficult to monitor or modify DNS requests. These changes could render an
organization’s security controls ineffective.

The Cybersecurity and Infrastructure Security Agency (CISA) recommends users
and administrators review the Dutch NCSC fact
sheet on DNS monitoring for additional information and recommendations.

The Microsoft Threat Intelligence Center (MSTIC) has released a blog post
describing an increase in malicious cyber activity from the Iranian group known
as Phosphorus. These threat actors are exploiting password reset or account
recovery features to take control of targeted email accounts.

The Cybersecurity and Infrastructure Security Agency (CISA) encourages users
to review the Microsoft
blog for additional information and recommendations and CISA’s Tip on Supplementing
Passwords.

    The Portable Document Format (PDF) standard has been able to provide many benefits that unify communications across many different software and hardware platforms. One of those elements is the encryption schemes that allow users to password protect their documents from view, edit, or saving permissions without the required password. Another encryption feature included with the PDF standard is the ability to sign documents with an electronic signature with the same legal standing as a handwritten signature, this may include digital signing which uses cryptographic measures to assure authenticity.

    Researchers from Ruhr University Bochum, FH Münster University of Applied Sciences, and Hackmanit GmbH have developed a two pronged attack on the security measures of PDFs and their encryption schemes. They have named their attack PDFex. In their research they developed methods for the exfiltration of the contents of the encrypted PDF with minimal prior knowledge of the contents of the PDF file. The methods studied can also modify the contents to change the plain text as well as add malicious functionality. The first prong of PDFex attack methods rely on how an encrypted PDF only encrypts portions of the PDF file leaving other portions unencrypted and unprotected. The attacker is then able to modify the contents of the unencrypted portions of the file. In this way they can plant data which submits a form including the contents of the PDF to an attacker controlled server granting the attacker access to the contents of the PDF. The attacker can edit an unencrypted field with a URL which will be sent encrypted and unencrypted strings from the document. The last method in this attack on the unencrypted portion of PDF files injects JavaScript code into the document which then ex filtrates the data within the file. This is the “Direct Exfiltration” method of the PDFex attack.

    The other prong of this attack uses CBC malleability gadgets, tools that are able to edit cipher texts encrypted with the cipher block chaining (CBC) encryption mode without integrity checks. It just so happens that the PDF standard does exactly that. This method can modify plain text as well as add in new encrypted content to the file. This technique can enact the PDF forms and hyperlink techniques as listed in the Direct Exfiltration method. The CBC Gadgets method can also edit PDF object streams such that they submit themselves to an attacker controlled server. Both attacks require the victim to open the tainted document so that the traps can deliver the finally decrypted information to the attacker. The researchers have tested their techniques on 27 PDF viewers and all were susceptible to at least one method of the PDFex attack.
    The attack requires that the attacker have access to the file to modify it, some of the attacks have other requirements such as the ability to trigger URL s from the viewer, or for the viewer to have permission to use JavaScript in the background. One of the researchers reported to Threatpost that “There are currently no effective countermeasures, as the weaknesses lie in the PDF encryption standard itself” and that the best mitigation is to use additional layers of encryption outside of the PDF standard to protect their data.

Sources:

• https://www.pdf-insecurity.org/download/paper-pdf_encryptionccs2019.pdf

• https://threatpost.com/hack-breakspdf-encryption/148834/

    Remote malware has been around for almost the entire history of computers. Attackers are always looking for ways to exfiltrate data from systems and be able to control their malware from a remote location. The Command & Control (C2) devices are usually servers controlled by the attacker, but a new malware dubbed by Juniper researchers as Masad has taken a different approach: using the messaging app Telegram for its C2 functions.

    Telegram is a popular messaging and Voice-over-IP (VoIP) app with over 200 million active monthly users. This makes it a pretty good place to try and hide malicious activity. Masad uses the sendDocument API of Telegram to exfiltrate data stolen from victims as a 7zip archive. Juniper has detected over 1,000 variants of Masad in the wild, as well as 338 unique Telegram C2 bots related to its use. Due to the malware being sold as a product rather than kept to a particular group, multiple groups can be using Masad for different campaigns. The developers of Masad have even created a group within Telegram with over 300 members, designed for potential clients and tech support.
Masad’s attack vectors include disguising itself as a legitimate tool or hiding itself in other third-party tools. For instance, it has been seen mimicking CCleaner, Utilman, Whoami, ProxySwitcher, a Samsung Galaxy software update, and many others. The developers have also included current trends in gaming, especially for younger internet users that may not be security conscious, by hiding Masad as Fortniteaimbot 2019.exe and an EXEA HACK CRACKED executable claiming to be for PUBG, CounterStrike Global Offensive, Fortnite, Grand Theft Auto 5, and DOTA.

    The malware also has the capability to download additional malicious tools, usually more cryptominers. Masad has a wide array of abilities for information stealing in addition to its cryptomining. It can steal system information including running processes, desktop files and screenshots, browser information such as cookies, passwords, credit cards, and AutoFill data, as well as Steam, FileZilla, and Discord files. Masad is also being advertised as a Clipper which looks for cryptocurrency wallet information in the system’s clipboard and replaces it with the attacker’s wallet information. It searches for over two dozen different flavors of cryptocurrency, including Bitcoin, Litecoin, Monero, Ethereum, and DogeCoin.

    Juniper researchers recommend locking down the Telegram communication protocol at the firewall level provided there is no legitimate business use that this would interrupt. They also suggest using a next-generation firewall with Advanced Persistent Threat (APT) protection to help counteract the malware if it gets inside the organization.

Sources

• https://threatpost.com/masad-spyware-telegram-bots/148759/ 

• https://coingeek.com/new-malware-uses-telegram-app-to-replace-cryptoaddresses/

• https://forums.juniper.net/t5/Threat-Research/Masad-Stealer-Exfiltratingusing-Telegram/ba-p/468559 

    Baseband Management Controllers (BMC) are a popular feature found on most motherboards targeting the server market. They provide a number of convenience functions for remote management which is great for machines typically located in a cold noisy room. Some of the functions they provide include remote power cycling, keyboard video mouse (KVM), and virtual media emulation. The combination of these functions can allow an administrator to provision a server without ever having to touch it. With that much power over the system they are bound to be a highly valuable target for attack.

    This week the security company Eclypsium released a critical vulnerability they found in Supermicro’s BMC implementation. The vulnerability reported is in the virtual media service subsystem. This service allows a remote administrator to attach USB devices, such as DVD drives or keyboards, to the machine remotely as if they were physically plugged into the machine. The feature requires authentication to function properly of course but the researchers found a way to bypass this requirement. 

    The first weakness is that the BMC would accept authentication requests via plaintext by default. They noted that encryption support is available but based on an old weak Rivest Cipher 4 algorithm. In addition, the key used when using encryption is shared across all Supermicro devices, making man-in-the-middle decryption possible. They also uncovered a complete authentication bypass in the system. This is possible because the BMC does not timeout a valid authorized session in a timely manner. An attacker would be able to re-use the session and gain access if an administrator had recently successfully logged into the system and used the virtual media service. BMC systems are rarely reset due to their nature of being an always online out of band management system, increasing the likelihood of this attack being successful.

    Supermicro has issued an update to their BMC software, but it is unlikely that machines will be patched immediately. This is due to the machines needing to be completely powered off in order to apply the update. Until then it is recommended to block the port used by the virtual media service, port 623, until the patch can be applied. Researchers warn that this will likely not be the last BMC vulnerability discovered, so additional measures should be taken when possible. The best defense against these attacks is keeping vulnerable machines on a separate network from other traffic. Ideally management interfaces should be on their own network that is not exposed to public facing traffic.

Sources

 • https://csoonline.com/article/3435900/insecure-virtual-usb-feature-insupermicro-bmcs-exposes-servers-to-attack.html

• https://eclypsium.com/2019/09/03/usbanywhere-bmc-vulnerability-opens-servers-to-remote-attack

    There has always been a battle between the Google Play Store and the malicious applications that attempt to reside on it. Google implements rigorous security testing of all apps, but some can still slip through the cracks. Such was the case when researchers from Symantec’s Threat Intelligence team found 25 instances of malicious apps, with a combined userbase of over 2.1 million, on the Google Play Store. These apps were designed to be camouflaged as photo utility and fashion apps, and upon download, did not exhibit any malicious properties. It wasn’t until the app downloads a remote configuration file that it becomes malicious. This behavior is what allows the app to bypass the security checks implemented by Google. Since the malicious code is not actually in the app and is downloaded remotely, Google is none the wiser. Researchers say that the 25 apps share a similar code structure, leading them to believe that the developers are part of the same organization or, at least, using the same code base. 

    Once installed, the app hides its icon and begins to display full-screen advertisements at random intervals with the app title hidden. This is done to prevent users from determining which app is responsible for the ads. This behavior continues even when the app is closed. This can be confusing for users who cannot even recall downloading the app as there is no icon or name associated with the behavior. Another interesting trick the developers use is the use of two versions of the same app. One version is a malicious version with full-screen advertisements while the other is a non-malicious version, which just so happens to be present in the Google Play’s Top App Charts. The researchers believe that this is done in the hope that users accidentally download the malicious copy of the app instead of the popular, non-malicious version. 

    The researchers believe that the primary reason for the creation of these apps is the monetary gain from the advertising revenue. There will be some subset of users that will continue to deal with the advertisements, despite their annoyance. When downloading apps from the Google Play Store, it can be difficult to determine which are malicious at first glance. In order to protect yourself from malicious applications, the researchers suggest keeping software updated, not downloading apps from unfamiliar sites, only installing apps from trusted sources, and noticing the permissions requested by apps that you download

Sources: 

• https://www.bleepingcomputer.com/news/security/malicious-androidapps-evade-google-play-protect-via-remote-commands/ 

• https://www.symantec.com/blogs/threat-intelligence/hidden-adwaregoogle-play09