ZeroHour
Recorded Futurepublished ()ingested CHRIS

Uncovering Hidden Lynx: Using OSINT for APT Analysis

Vulnerabilities mentionedAll →

CVEVulnerabilityCVSSEPSSFlagsAffectedExposurePublished
CVE-2012-1889
Memory Corruption RCE in Microsoft XML Core Services

Microsoft XML Core Services (MSXML) contains a memory corruption flaw (CWE-119) that can allow remote code execution when the component processes attacker-crafted content, such as a malicious web page or XML document loaded through Internet Explorer. Successful exploitation lets an attacker run arbitrary code in the security context of the logged-on user, gaining the ability to install programs, and view, change, or delete data; an admin-context victim would yield full system compromise. Any Windows system shipping the affected MSXML components is affected, which at the time of disclosure meant essentially the entire Windows installed base. Exploitation is confirmed in the wild: CISA added the flaw to the Known Exploited Vulnerabilities catalog on 2022-06-08, EPSS assigns it an 83.6% probability of exploitation in the next 30 days (top percentile), and related reporting ties it to observed in-the-wild and APT activity, though public ransomware association is listed as unknown. No public proof-of-concept is cataloged, but the KEV and EPSS signals indicate active, ongoing exploitation risk.

Do: Apply the vendor updates for XML Core Services released in the July 2012 Patch Tuesday on all Windows systems, prioritizing internet-facing and legacy endpoints, per the CISA KEV required action. Because exploitation is triggered via crafted web/XML content, restrict or upgrade legacy Internet Explorer usage and block untrusted web content on affected hosts. Hunt for exploitation indicators given the 2022-06-08 KEV listing and high EPSS score, and verify MSXML-related patches are present in your patch inventory.

84% KEV
  • Microsoft XML Core Services (MSXML)
mass≈hundreds of millions of Windows endpoints (MSXML ships as a core Windows component)
CVE-2013-3893
Memory Corruption RCE in Microsoft Internet Explorer

CVE-2013-3893 is a resource-management (memory corruption) flaw in Microsoft Internet Explorer that can allow remote code execution (CWE-399). It is triggered remotely, typically when a user views attacker-controlled web content in a vulnerable version of Internet Explorer. A successful attacker gains the ability to execute arbitrary code in the context of the current user, potentially compromising the workstation. Organizations still running Internet Explorer, which CISA notes may be end-of-life (EoL) and/or end-of-service (EoS), are affected; specific affected version ranges were not provided in the source data. The flaw was patched in Microsoft's October 2013 Patch Tuesday after being exploited in the wild (Operation DeputyDog, per related reporting), and CISA added it to the Known Exploited Vulnerabilities catalog on 2025-08-12 with a very high EPSS of 85.9% (100th percentile), indicating active or imminent exploitation.

Do: Apply mitigations per Microsoft's vendor instructions and follow applicable BOD 22-01 guidance for cloud services, or discontinue use of Internet Explorer if mitigations are unavailable, per CISA's required action. Verify that affected systems have the October 2013 Patch Tuesday (or later) cumulative Internet Explorer security updates installed, and audit your estate for remaining legacy IE usage. Where IE is still needed for legacy sites, migrate to Microsoft Edge with IE mode and treat in-the-wild exploitation as likely given the KEV listing and 85.9% EPSS.

86% KEV
  • Microsoft Internet Explorer
masstens to hundreds of millions of legacy Windows devices historically capable of running IE; current actively used legacy IE installs unknown but plausibly in…
Full article1,289 words · extracted from recordedfuture.com · click to collapse

Analysis Summary

  • Visualization of open source intelligence on APTs reveals overlapping infrastructure, tools, and exploits used in the VOHO campaign and Operations Aurora, DeputyDog, and Ephemeral Hydra.
  • Two vulnerabilities were identified as exploited by Hidden Lynx in its VOHO campaign (2012) and the Elderwood Gang responsible for Operation Aurora (2010). Command and control infrastructure was also shared between Hidden Lynx and threat actors responsible for two campaigns during 2013: Operation DeputyDog and Operation Ephemeral Hydra.
  • Threat intelligence derived from disparate open web sources bolsters security efforts by identifying and contextualizing links between threat actors.

When the_New York Times_ and Mandiant last year unmasked a large scale Chinese hacking operation, pinpointing its location down to the building, the report drew mainstream attention to what security professionals already well knew: sophisticated threat actors carry out persistent cyber operations over months and years.

This post illustrates how open source intelligence (OSINT) from the web is not only useful but crucial to threat intelligence teams analyzing advanced persistent threats. Through OSINT analysis, we discover clues to APT attack methods that link together the research of independent threat intelligence teams, suggesting new lines of analysis that were not prominent when looking at the work of any single threat intel team in isolation.

We will detail how the C&C infrastructure and tools used by hacker group Hidden Lynx during its VOHO campaign (2012), excellently documented by Symantec researchers last September, overlap with tools used in other high profile operations during the past few years. The other campaigns evaluated are Operation Aurora (2010), Operation DeputyDog (2013), and Ephemeral Hydra (2013).

History of Hidden Lynx Malware

Security researchers have shown time and again that cyber actors frequently recycle effective attack methods. These attack methods retain telltale signatures even as a threat actor’s targeting and tools evolve. Therefore, our investigation looked for clues that linked hackers across one or more of the campaigns in question.

Seeking to tease out any possible links between Operation Aurora, VOHO, Operation DeputyDog, and Ephemeral Hydra, we began with Symantec’s Hidden Lynx report as our foundation. The authors of that report identify three primary tools used in the campaigns attributed to Hidden Lynx: Trojan.Naid, Backdoor.Moudoor, and Backdoor.Hikit.

Using Recorded Future, we quickly built a timeline of the reported use of those tools in major security incidents, finding many events prior to the early 2013 exposé on Hidden Lynx.

Click image for larger view

Connecting the Dots

Our analysis, based solely on information openly available on the web, builds on significant technical research that has explicitly linked elements of the aforementioned campaigns. In particular, FireEye during the fall of 2013 called out infrastructure overlap between Ephemeral Hydra and DeputyDog.

Aspects of these two campaigns used tools and C&C infrastructure reportedly used by Hidden Lynx. These include a Hydraq/McRAT variant with lineage traceable to Aurora. Symantec during 2012 linked the Elderwood Project to Operation Aurora; Trojan.Naid and Backdoor.Moudoor were also used in Aurora, by the Elderwood Gang, and by Hidden Lynx.

A brief summary of targets for each campaign:

  • Aurora: Technology, Finance, Energy and Defense
  • Elderwood: Defense and Supply Chain
  • VOHO: Government, Defense, and Education (Hidden Lynx separately targeted Finance)
  • DeputyDog: Government (Japan and China)
  • Ephemeral Hydra: Government and Defense

Much of the above was well documented. However, some connections described over the course of several years, oftentimes by different research teams published across a variety of sources, are difficult to recall or might even be obscured by more recent information.

What links did we discover in analysis using Recorded Future?

  • Two vulnerabilities were found to be exploited by both the Elderwood Gang and Hidden Lynx: CVE-2012-1875 and CVE-2012-1889.
  • One of those common exploits, targeting CVE-2012-1889, was on November 22, 2012 reportedly added to the “popular” Gong Da exploit kit.
  • Three C&C servers were known to be shared between VOHO and various elements of Operation DeputyDog and Operation Ephemeral Hydra: 58[.]64[.]143[.]244, 66[.]153[.]86[.]14, 111[.]68[.]9[.]93
  • Backdoor.Hikit, which appears to have been developed post-Operation Aurora, was used in both VOHO and DeputyDog campaigns.
  • Reported use of the Hydraq Trojan from Aurora (2010) fell off until September 2012. The delivery method appears to have then switched from spearphishing to a watering hole attack when a variant of Hydraq was discovered in Ephemeral Hydra during November 2013.
Click image for larger view

The above network shows relationships between three tools used by Hidden Lynx during its VOHO campaign: Trojan.Naid, Backdoor.Moudoor, and Backdoor.Hikit. The connections indicate actors, targets, vulnerabilities, and technical information related to those tools. Note the density of connections between Moudoor and Naid, which are commonly reported as complementary tools.

Trojan.Naid specifically was used to attack security firm Bit9 ultimately compromising its digital code signing certificate. This allowed the hackers to then take on higher value targets as part of the VOHO campaign. Naid was also used in the 2010 campaign known as Operation Aurora.

One of the key findings noted above was the identification of two vulnerabilities that were exploited by both Hidden Lynx and the Elderwood Gang, which has been linked to Operation Aurora. CVE-2012-1875 and CVE-2012-1889 are at the center of the network below, surrounded by targets and other related vulnerabilities.

Click image for larger view

After visualizing these links, we explored the vulnerabilities at the edges of the graph, looking for connections with exploits used in other campaigns. We found such a link in CVE-2013-3893, which was leveraged in the August 2013 attacks on Japanese targets that FireEye dubbed Operation DeputyDog.

Similar to our approach with Symantec’s report on Hidden Lynx, we used Recorded Future to organize the technical details about the DeputyDog attacks to reveal technical information described in the open source reporting across multiple campaigns.

Click image for larger view

The above network shows technical details and locations, both server hosts and targets, related to Operation DeputyDog. This enabled creation of useful lists for us to seek overlap with other operations, establish the timing of exploits and C&C infrastructure used, and arrange alerts for ongoing monitoring.

To wrap up our investigation, we built a timeline of events related to the IP addresses associated with both Hidden Lynx campaigns as well as operations DeputyDog and Ephemeral Hydra. The timeline details events related to the three previously cited IPs – 58.65[.]143[.]244, 66[.]153[.]86[.]14, and 111[.]68[.]9[.]93 – used in both the Bit9 attacks and either DeputyDog or Ephemeral Hydra.

Analysts were able to track 66[.]153[.]86[.]14 back to an email address linking DeputyDog and the Hidden Lynx VOHO campaign. Separate analyses in November 2013 describe C&C servers used by Trojan.APT.9002 also used in the Hidden Lynx attacks against Bit9.

Click image for larger view

Revealing Links and Context With Web Intelligence

This investigation sheds new light on these campaigns, without depending on private or freshly disclosed technical intelligence. Once assembled and structured for analysis, the collective knowledge about Hidden Lynx and many other APT campaigns, some publicly available for years, allows us to better understand intent and capabilities.

We often hear leaders lament the present state of intelligence sharing, and advocate for new approaches with hold great promise. Meanwhile, large amounts of intelligence are already being shared in open source reporting. To yield insights, we must deal with great variety in sources and formats. The above data points were derived from PDFs, blog posts, and write ups from security firms coming from hundreds of websites. Recorded Future captures and structures that information, and provides analytic tools to help quickly make sense of this information from disparate sources.

The results of such threat actor and campaign analysis, corroborating evidence across multiple sources and linking technical details across security blogs, social media, and research forums, are insights that are useful from incident response to threat intelligence.

Text extracted automatically; images, tables and formatting may be missing. Original: https://www.recordedfuture.com/blog/hidden-lynx-analysis