What is malware that was never written? Ask Webworm. The China-aligned espionage crew has been caught spying on government networks with a toolkit that contains no custom malware at all, just a doctored copy of a common, legit VPN tool.
“For a threat actor, the perfect malware is one they don’t have to write, and already does all they need,” Rapid7 said in its report on the operation.
Rapid7 ties the tool to Webworm with moderate confidence, and says the group’s targets are government organizations, particularly in Europe.
“One example of this is shown by the China-aligned APT known as Webworm, which has been known to deploy the open-source SoftEther VPN into the networks of compromised governmental organizations to maintain low-profile, long-term access.”
The crew has been on this path for a while. In May, ESET reported that starting in 2024, Webworm began moving away from traditional backdoors toward legitimate and semi-legitimate tools like SoftEther. That same year, its targeting shifted from Asia to Europe, and in 2025 ESET saw attacks on government organizations in Belgium, Italy, Serbia and Poland. The enlightenment is partial: ESET also caught the group building new custom backdoors in 2025. The latest build takes the no-code approach one step further.
Webworm Practices Wu Wei: Hacking by Not Hacking
The tool, which the attackers call NtMonitor, is a modified build of the SoftEther VPN Bridge. It blends in as a normal Windows file, posing as “Network Monitor Software” from a “Network Systems Solutions Inc.” under the filename netcapa.exe. A VPN that does not call itself a VPN.
[Related: Free VPN Management Tool for Criminals Is Itself a Backdoor]
The disguise isn’t flawless. Some internal strings still read “SoftEther VPN,” the service is still named SEVPNBRIDGE, and many registry paths still point to the “SoftEther Project.” But the attackers changed the details a defender would look for first. The stock web endpoints, /vpnsvc/vpn.cgi and /vpnsvc/connect.cgi, became /update/alive.cgi and /update/picture.cgi. The config file was renamed setting.conf and encrypted with a custom RC4-like cipher. (Rapid7 has published a script to decrypt it.)
Then there’s the handshake.
“NtMonitor changes the upstream GIF watermark with a custom byte pattern (a truncated PNG file showing the Windows logo), so that stock SoftEther clients and servers cannot handshake with this build, and NtMonitor’s own peers cannot be fingerprinted with the well-known stock GIF watermark,” the Rapid7 team explained.
The Path of Least Detection
While simple, the technique has proven to be highly effective. The sample has been publicly available since January. When Rapid7 analyzed it, exactly one antivirus vendor on VirusTotal flagged it.
Once in place, the bridge joins the infected machine’s local network to the attackers’ own, and the compromised host serves as a proxy the attackers can use to reach other machines inside the network. No backdoor, no payload. Beyond the disguise, Rapid7 says it found no additional malware functionality, and it checked the build against SoftEther’s public source code to make sure.
“Clearly, the threat actor prioritizes stealth above raw capability,” the researchers note.
“It seems like they methodically analyzed SoftEther’s detection surface and implemented counter-measures to all of the easy ways a detection engineer may target it.”
Finding the Way Back
Rapid7 has posted indicators of compromise and YARA rules on its Rapid7-Labs GitHub repository. For network defenders with TLS inspection, the most reliable tell is a POST request to /update/picture.cgi, labeled as a JPEG image, whose body actually starts with a PNG header.
Rapid7 warns that those endpoints and the watermark are easy to change. Its advice to government defenders is to learn what normal SoftEther traffic looks like on their networks.
Know the Way, and you’ll notice who’s wandering off it.