LV Ransomware

Threats & Defenses

Summary

Secureworks® Counter Threat Unit™ (CTU) researchers investigated reports that the LV ransomware had the same code structure as REvil. This overlap could indicate that the GOLD SOUTHFIELD cybercriminal threat group that operates REvil sold the source code, that the source code was stolen, or that GOLD SOUTHFIELD shared the code with another threat group as part of a partnership. CTU™ analysis confirmed that the GOLD NORTHFIELD threat group, which operates LV, replaced the configuration of a REvil v2.03 beta version to repurpose the REvil binary for the LV ransomware.

CTU researchers have not observed LV ransomware advertisements on underground forums as of this publication. However, variations in partner and campaign IDs across LV configurations and the practice of naming and shaming victims could indicate that GOLD NORTHFIELD is launching a ransomware-as-a-service (RaaS) offering.

LV packer

The packed LV ransomware samples identified by CTU researchers appear to use the same basic crypter. Figure 1 shows the entire contents of the packed executable’s main function, which contains five of the executable’s nine functions.

Main function for the packer used to unpack and execute LV ransomware.
Figure 1. Main function for the packer used to unpack and execute LV ransomware. (Source: Secureworks)

The packed executable stores the LV ransomware binary as RC4-encrypted data within a section named ‘enc’ (see Figure 2).

Encrypted LV ransomware binary stored in the enc section of the packer executable.
Figure 2. Encrypted LV ransomware binary stored in the enc section of the packer executable. (Source: Secureworks)

The packed samples analyzed by CTU researchers use the hard-coded “kZlXjn3o373483wb6ne1LIBNWD3KWBEK” key to decrypt the contents of the enc section. The “This program cannot be run in DOS mode” string is removed from decrypted binaries’ PE headers (see Figure 3).

Strings removed from header of decrypted LV binary.
Figure 3. Strings removed from header of decrypted LV binary. (Source: Secureworks)

Once decrypted, the ransomware binary is copied into a new memory allocation created with Read/Write/Execute (RWE) access rights. The packer then executes the unpacked ransomware binary by jumping to the entry point defined in the PE header.

Origin and configuration

The code structure and functionality of the LV ransomware sample analyzed by CTU researchers are identical to REvil. The version value in the LV binary is 2.02, its compile timestamp is 2020-06-15 16:24:05, and its configuration is stored in a section named ‘.7tdlvx’. These characteristics align with REvil 2.02 samples first identified in the wild on June 17, 2020. The LV sample also contains a code segment that is unique to REvil 2.03. The only purpose of this code segment in REvil binary is to taunt prominent security researchers. LV replaces the insults with the space character (see Figure 4).

LV code segment duplicating REvil 2.03 code but replacing strings with spaces.
Figure 4. LV code segment duplicating REvil 2.03 code but replacing strings with spaces. (Source: Secureworks)

This type of code modification suggests that GOLD NORTHFIELD does not have access to REvil’s source code. The threat actors likely used a hex editor to remove potentially identifying characteristics from the binary to conceal that LV is a repurposed version of REvil. The hard-coded 2.02 version value and the unique REvil 2.03 code suggests that GOLD NORTHFIELD used a beta version of REvil 2.03 as the basis for LV ransomware.

REvil binary repurposing

The GOLD NORTHFIELD threat actors replaced the REvil configuration stored within the binary’s ‘.7tdlvx’ section with their own configuration. Successful replacement required the format of the REvil and LV configurations to be identical. Figure 5 shows the LV configuration extracted from the REvil binary. It is a JSON-formatted string containing key elements utilized by REvil (e.g., sub, net, dmn, pk).

LV configuration.
Figure 5. LV configuration. (Source: Secureworks)

GOLD NORTHFIELD then had to RC4-encrypt the LV configuration with a 32-byte key. To bypass REvil’s anti-tamper control that ensures the integrity of the configuration (see Figure 6), GOLD NORTHFIELD also had to generate a CRC32 hash of the updated encrypted configuration and then replace the hard-coded precalculated CRC32 hash stored in the binary with the updated configuration’s CRC32 hash. These changes are necessary because the REvil code calculates the configuration’s CRC32 hash value at runtime and terminates if the calculated and hard-coded hashes do not match.

Configuration anti-tamper control implemented in the REvil binary.
Figure 6. Configuration anti-tamper control implemented in the REvil binary. (Source: Secureworks)

Finally, GOLD NORTHFIELD could add the RC4 key used to encrypt the configuration, the CRC32 hash of the encrypted configuration, the length of the encrypted configuration, and the encrypted configuration itself to the REvil binary via the identified configuration section (.7tdlvx) in the defined order (see Figure 7).

REvil configuration structure when stored in the binary.
Figure 7. REvil configuration structure when stored in the binary. (Source: Secureworks)

If done correctly, the binary will successfully execute using LV’s updated configuration. Files on the victim’s system will be encrypted with session keys that are protected by LV’s public key, and victims will be directed to LV’s ransom payment site via the updated ransom note.

Configuration comparisons

CTU analysis of numerous LV configurations led to several insights:

Ransom payment site

After accessing the ransom payment site, victims are presented with a basic form that requests the key from the ransom note (see Figure 9).

LV ransom payment site key submission form.
Figure 9. LV ransom payment site key submission form. (Source: Secureworks)

Previous CTU analysis of the REvil ransom note determined that this key represents information about the ransomware infection that has been encrypted and then Base64-encoded:

  • Compromised host details:
    • CPU architecture (32-bit or 64-bit)
    • Fixed-drive information (drive letter, drive type, total size, and free space)
    • Workgroup/domain
    • Configured locale, and whether it aligns with one of the specified countries where the malware cannot be used
    • Hostname
    • Operating system
  • Ransomware details:
    • Configured partner ID
    • Threat actor’s configured public key
    • Encrypted session private key
    • Configured campaign ID
    • Unique ID based on host’s volume serial number and CPUID
    • Victim’s username
    • Ransomware version

As of this publication, CTU researchers have identified three ransom payment Tor domains specified in LV ransom notes. Each of the domains successfully loads the landing page, but CTU researchers’ attempts to submit the key from the ransom note returned HTTP errors (see Table 1).

Ransom payment domain HTTP error
4to43yp4mng2gdc3jgnep5bt7lkhqvjqiritbv4x2ebj3qun7wz4y2id . onion 502 – Bad Gateway
l55ysq5qjpin2vq23ul3gc3h62vp4wvenl7ov6fcn65vir7kc7gb5fyd . onion 403 – Forbidden
36yvrbzhbzyuzia7qxahsaw2yizcr3heljw2jtde3smyuhkokjnb2sid . onion 403 – Forbidden

Table 1. LV ransom payment domains and HTTP errors.

The HTTP errors may be caused by anti-analysis controls implemented by GOLD NORTHFIELD to inspect characteristics of the submitted key for suspicious or undesirable activity. They may also indicate that the threat group is struggling to maintain resilient infrastructure due to lack of skill or insufficient resources.

When key submission is successful, the site displays a page showing the ransom amount in U.S. dollars and how much time the victim has to pay the ransom before sensitive data is disclosed (see Figure 10). The page also includes a live chat function for the victim to interact with the threat actors.

LV ransom payment site.
Figure 10. LV ransom payment site. (Source: ID Ransomware blog)

Leak sites

CTU researchers identified two LV ransomware leak sites that have an identical structure but appear to be operated independently. The sites listed victims during the same timeframe, but only one victim was listed on both sites (see Figure 11). It is unclear why GOLD NORTHFIELD would operate two leak sites.

Victims added to LV leak sites between March 7 and April 14, 2021. Only one victim (highlighted in yellow) was listed on both sites.
Figure 11. Victims added to LV leak sites between March 7 and April 14, 2021. Only one victim (highlighted in yellow) was listed on both sites. (Source: Secureworks)

The leak sites name and shame victims. The threat actors coerce the victims into paying the ransom by threatening to publish their sensitive information (see Figure 12).

LV leak site.
Figure 12. LV leak site. (Source: Secureworks)

GOLD NORTHFIELD typically threatens to publicly release sensitive information if victims do not initiate contact within 72 hours. The threat actors post screenshots of the victim’s sensitive files on the leak sites to support their claims. However, it appears that none of the victims’ data has been released as of this publication. It is unclear if victims paid the ransom and the threat actors just keep the full list of victims on the leak site as evidence of their conquests.

Conclusion

CTU analysis revealed that the LV ransomware is not a distinct ransomware family; it is repurposed REvil ransomware. By modifying the binary of a prolific ransomware family, the GOLD NORTHFIELD threat actors significantly expedited their maturity within the ransomware ecosystem. Without expending resources on ransomware development, the group can operate more efficiently than its competitors while still offering a best-in-class ransomware offering, ultimately resulting in a more profitable business model. GOLD NORTHFIELD’s unauthorized manipulation of REvil will likely prompt GOLD SOUTHFIELD to implement additional anti-tamper controls and modify configuration storage and processing to impede future attempts to overwrite the REvil configuration.

It is too early in GOLD NORTHFIELD’s evolution to evaluate the threat it poses. The ability to repurpose the REvil binary suggests that the threat actors have technical capabilities. Additionally, the complexity required for this repurposing and the configuration variations across LV samples suggest that GOLD NORTHFIELD may have automated the process. Although a RaaS for the LV ransomware could provide direct competition for GOLD SOUTHFIELD’s RaaS offering, the lack of a reliable and organized infrastructure needed to operate a successful RaaS offering suggests that GOLD NORTHFIELD has to expand its capabilities and resources to compete with other ransomware operations.

Threat indicators

The threat indicators in Table 2 can be used to detect activity related to LV ransomware. The domains may contain malicious content, so consider the risks before opening them in a browser.

Indicator Type Context
6f0b92488eae3ccefc0db7a6b0d652ee MD5 hash Packed LV ransomware
45adc4224d2ae9fd75b19417ca6913515c5222ee SHA1 hash Packed LV ransomware
457936c28938616495836c472b3389a0870574bee6
a5dc026d5bd14979c6202c
SHA256 hash Packed LV ransomware
58682ca2a49ed4bfb8d5aaf76cf0fade MD5 hash Packed LV ransomware
b00d58e9ffd784db86e77a6a31c76e1bd58ba79b SHA1 hash Packed LV ransomware
ab2f84103e95806b25c6d163d6210a21fb3283cd29
dddee917d33e654d733425
SHA256 hash Packed LV ransomware
7b1cf5fc0bfb1021fe0e14e518c32026 MD5 hash Packed LV ransomware
380cd990a9e5aec85233ef1d2635dc04d5a96e6b SHA1 hash Packed LV ransomware
d4fc76bf8baae39feec23990857c52199e80265a34
faeece0d830eb77645c944
SHA256 hash Packed LV ransomware
a4331ff805b0a8f2a2892777c224b65e MD5 hash Packed LV ransomware
2c5521077dd1a6f5f3558351370880aee9ab7c71 SHA1 hash Packed LV ransomware
329983dc2a23bd951b24780947cb9a6ae3fb80d5ef
546e8538dfd9459b176483
SHA256 hash Packed LV ransomware
fa8117afd2dbd20513522f2f8e991262 MD5 hash Packed LV ransomware
f7b876edb8fc0c83fd8b665d3c5a1050d4396302 SHA1 hash Packed LV ransomware
78b592a2710d81fa91235b445f674ee804db39c8cc
34f7e894b4e7b7f6eacaff
SHA256 hash Packed LV ransomware
d1c9c12e08c8e2111da989e2318b1c42 MD5 hash Unpacked LV ransomware
d0c7f3c8de28d0fccec9d4925afeb5fa9dd62b5d SHA1 hash Unpacked LV ransomware
e25eaaac03aa958688cbe950275156169eb4955e14
5bc9627fcbfb36cd832a84
SHA256 hash Unpacked LV ransomware
4to43yp4mng2gdc3jgnep5bt7lkhqvjqiritbv4x2e
bj3qun7wz4y2id.onion
Domain name LV ransomware payment site
l55ysq5qjpin2vq23ul3gc3h62vp4wvenl7ov6fcn6
5vir7kc7gb5fyd.onion
Domain name LV ransomware payment site
36yvrbzhbzyuzia7qxahsaw2yizcr3heljw2jtde3s
myuhkokjnb2sid.onion
Domain name LV ransomware payment site
rbvuetuneohce3ouxjlbxtimyyxokb4btncxjbo44f
bgxqy7tskinwad.onion
Domain name LV ransomware leak site
4qbxi3i2oqmyzxsjg4fwe4aly3xkped52gq5orp6ef
pkeskvchqe27id.onion
Domain name LV ransomware leak site

Table 2. Indicators for this threat.

References

Gillespie, Michael (@demonslay335). “Sodinokibi.” Twitter, November 13, 2020, 5:24 pm. https://twitter.com/demonslay335/status/1327376936935493635

Ivanov, Andrew. “LV Ransomware.” ID Ransomware blog. November 13, 2020. Original (in Russian): https://id-ransomware.blogspot.com/2020/10/lv-ransomware.html. English translation: https://translate.google.com/translate?hl=en&sl=ru&u=https://id-ransomware.blogspot.com/2020/10/lv-ransomware.html&prev=search&pto=aue

Secureworks. “GOLD NORTHFIELD.” Accessed June 1, 2021. https://www.secureworks.com/research/threat-profiles/gold-northfield

Secureworks. “GOLD SOUTHFIELD.” Accessed April 22, 2021. https://www.secureworks.com/research/threat-profiles/gold-southfield

Secureworks. “REvil/Sodinokibi Ransomware.” September 24, 2019. https://www.secureworks.com/research/revil-sodinokibi-ransomware

xiaopao (@Kangxiaopao). “LV ransomware.” Twitter, October 23, 2020, 5:43 am. https://twitter.com/Kangxiaopao/status/1319575086995652609

Products You May Like

Articles You May Like

DeceptionAds Delivers 1M+ Daily Impressions via 3,000 Sites, Fake CAPTCHA Pages
US Government Issues Cloud Security Requirements for Federal Agencies
Thai Officials Targeted in Yokai Backdoor Campaign Using DLL Side-Loading Techniques
US Organizations Still Using Kaspersky Products Despite Ban
Italy’s Data Protection Watchdog Issues €15m Fine to OpenAI Over ChatGPT Probe

Leave a Reply

Your email address will not be published. Required fields are marked *