Rather than building new malware from scratch, a China-aligned espionage group tracked as Webworm appears to have taken a simpler route back into compromised government networks: repurposing a legitimate, open-source VPN tool and giving it just enough of a disguise to slip past defenders expecting something more obviously malicious.
A Bridge, Rebranded
Researchers at Rapid7 described the find in a November 19 report, attributing it to Webworm with moderate confidence. The tool in question is a modified version of SoftEther VPN Bridge — open-source software normally used to link two networks together — internally labeled “NtMonitor VPN” in some of its strings and configuration data.
Comparing a publicly available 64-bit Windows sample against the matching stock SoftEther source release (version 4.44, build 9807), Rapid7’s analysts found no added command-execution backdoor baked into the code. That’s a meaningful distinction from Webworm’s previously documented GraphWorm backdoor, which communicated through Microsoft OneDrive and gave operators a command shell. NtMonitor serves a narrower purpose: it establishes a network route between an external system and the victim’s internal network, which on its own can be enough to reach other machines once a foothold exists.
Notably, at the time Rapid7 analyzed it, only a single antivirus vendor on VirusTotal flagged the sample — despite the file having been publicly available since January.
A Disguise With Loose Threads
The modified build presents itself as “Network Monitor Software” from a company called “Network Systems Solutions Inc.,” with an original filename of netcapa.exe. The cover story isn’t airtight, though: the underlying Windows service still registers under the name SEVPNBRIDGE, and several registry paths continue to reference the SoftEther Project by name. The build also bundles a resource archive that SoftEther normally keeps separate and stores its configuration in an encrypted file — changes that can make an installation look unfamiliar even to a team that knows what a standard SoftEther deployment should look like, without fully erasing the giveaways underneath.
The group also altered the software’s network fingerprint. Rapid7 found web request paths dressed up to resemble routine software-update traffic, along with a modified marker in the initial handshake between VPN peers — changes that would render detection signatures written for stock SoftEther traffic considerably less effective.
What’s Confirmed, and What Isn’t
It’s worth being precise about the limits of this report. Rapid7’s findings explain how Webworm maintains access after a breach — not how the group gets in to begin with, which agencies may have been affected by this specific tool, or what data, if any, attackers obtained through it. The researchers also found overlap between non-test configurations and IP addresses in AS20473, a hosting network associated with Vultr and The Constant Company, which lines up with infrastructure patterns seen in earlier Webworm activity — though the hosting network alone isn’t proof of compromise. Separately, ESET’s earlier research on Webworm documented past targeting in Belgium, Italy, Poland, Serbia, and Spain, but that describes prior activity rather than a confirmed victim list tied to NtMonitor specifically.
Detection Guidance for Government IT Teams
Because the risk here stems from an existing, legitimate feature being used without authorization rather than a new software vulnerability, the most useful starting point for defenders is a clear inventory: which VPN products are approved for use, which hosts are permitted to run them, and which remote networks they’re supposed to connect to. Investigating unexpected Windows services and firewall rule changes alongside connections to unapproved network peers — rather than any one signal in isolation — is likely to surface this kind of activity faster than relying on antivirus detection alone, given how poorly the sample was flagged in testing.
If an unauthorized bridge is found, incident responders are advised to preserve the executable, its configuration file, and connection logs before removing it, since that evidence is what allows a team to reconstruct which systems the access route reached — information a single compromised host rarely reveals on its own.
Leave a Reply
You must be logged in to post a comment.