Entities

Entities collected from this article, grouped for quick reference and copying. Inclusion does not establish malicious activity. Article links appear separately at the bottom as references.

A China-nexus activity cluster compromised a piece of software rarely associated with cyber attacks. UNC3569, an activity cluster tracked by Google, targeted the Sogou input method editor (IME), one of the most widely used Chinese-typing tools on Windows and owned by Tencent.

The flaw, CVE-2026-51990, opened a path to remote code execution through the embedded browser inside Sogou. In the attack Gen observed, UNC3569 chained it with an old Chromium V8 flaw. One crafted link and one click let the threat actor do essentially anything the logged-in user could do.

An IME is not an app anyone opens and closes. It sits resident in the background, running constantly and processing typed text, because that is its entire job. It holds deep, persistent, trusted access to the system by design. And almost nobody, neither users nor most security teams, thinks of it as attackable software. That is exactly what makes it a beautiful target. The more invisible and trusted a component is, the less anyone watches it.

In this instance, the threat actor used GRAYRABBIT as the payload, a backdoor UNC3569 has leaned on for years. It opens a remote shell, moves files in and out, and pulls down more tools on command.

GRAYRABBIT talks to its server on port 443, the port normally used for TLS web traffic. This time, however, the traffic is not encrypted the standard way. It is RC4-scrambled raw TCP riding on port 443 instead of TLS. Non-TLS traffic on port 443 is a quiet tell something is wrong, and defenders can hunt for it.

To the credit of Tencent, the response was fast. The flaw was reported April 9, then fixed and auto-pushed to all users by April 21. Twelve days. That is how responsible disclosure is supposed to work, and it mostly closed the door.

The room behind the door is still ugly. Gen found the embedded browser remained Chromium 80, still ran without a sandbox, and still had major browser security controls disabled after the fix.

The GRAYRABBIT loader self-deletes, so the malicious file may already be gone from the disk of a compromised machine, making it harder to know whether it was hit. No one has confirmed whether installing the patch removes a backdoor already running. Tencent never said which versions were vulnerable or how to check an installed one. The fix stops new infections, but anyone compromised before April 21 has no clear way to know it, or to clean it up.

Attention goes to guarding the doors everyone can see. Apps like the browser, the email client, and the productivity suite get the scrutiny. But in rare cases like this, the initial access vector was software so mundane nobody thought to look at it. An input method editor is trusted, invisible, and everywhere, and the combination is precisely what a threat actor salivates over. Trusted software nobody thinks about is software nobody defends.

The most dangerous software on a machine may not be the thing someone installed to do something. It may be the thing everyone forgot was there, trusted for so long nobody remembers to question it. Trust is not a security control. Sometimes it is the vulnerability.