ShadowCloud spyware campaign hides in GitHub and SharePoint traffic
Kaspersky says ShadowCloud spies on government networks in Asia and Africa by hiding commands inside GitHub, SharePoint and Ably traffic that looks perfectly normal.

Kaspersky has uncovered a spying campaign called ShadowCloud, and its main trick is blending in. The attackers pass their commands through services most offices use every day: GitHub, Microsoft SharePoint and a messaging service called Ably.
According to the company, its Global Research and Analysis Team (GReAT) found the campaign aimed at government bodies in Southeast Asia, Africa and Central Asia. It is still active and targets Windows computers. The researchers think the attackers got in through an earlier infection on those machines.
How does ShadowCloud work?
Once inside, the attackers use MeshCentral, a legitimate remote administration tool that gives access to the desktop, command line and files. It connects to the attackers' server and drops three files on the victim's computer: a genuine program, a malicious DLL and an encrypted malware file.
The genuine program is used to load the bad DLL, a technique called DLL side-loading. DLLs are software libraries Windows normally uses to load extra functions. The malicious one then decrypts and runs the payload, which activates a backdoor called GitDoor or GraphDoor. Kaspersky says neither had been seen before. Both collect system information and talk to GitHub, SharePoint and Ably to send data out and receive orders.
What can the attackers do?
Quite a lot. Kaspersky lists keystroke logging, screenshots, and searching, moving, uploading and downloading files. They can also run commands, manage Windows services, change network connections and firewall rules, tamper with processes and the system registry, and pull credentials out of memory. The malware even looks for the keys that decrypt WhatsApp databases.
The researchers found signs of possible links to Chinese-speaking threat groups, including FamousSparrow and CopperTurtle. They describe those as evidence of a connection, not a confirmed attribution.
Sourabh Sharma, principal security researcher at GReAT, says the techniques may look familiar but are hard for traditional network and file scanning to catch. His advice to organisations: watch remote administration tools and cloud API activity, look for unusual DLL loading and in-memory execution, and don't rely only on classic network controls and antivirus.
Apps worth a look
Tap the right tiles, add them up, hit the target. Easy to learn, hard to master. Free, for all ages.
Talk to an AI conversation partner; every sentence comes back corrected and explained in your language. 9 languages, A1 to C2.
Open EDIFACT and X12 messages, see every segment explained, convert to JSON/XML, export PDF and Excel. Runs locally, no upload.
Pick who it is for, the occasion and their personality. The search engine reads product descriptions from local shops and ranks what fits. Free, no sign-up.
More from this day
See all →
Uber data breach: 149,346 people hit via Fireworks AI hack

Blinkit barred from using CSK superfan's likeness in ads

Jensen Huang to appear before US House committees on December 1



