Cybercriminals Target AI Developers with Trojanized GitHub Repositories

Cybercriminals have shifted their tactics to target AI developers by cloning popular GitHub repositories for AI tools and developer resources. According to Netskope Threat Labs, these malicious repositories are part of a campaign that distributes an infostealer via trojanized code.

The attackers clone well-known repositories and subtly integrate malicious payloads, often adding the payload to a benign-looking subdirectory or modifying the URLs on the installation instructions. This makes it difficult for developers to detect the malware.

The 'lures' found among the impersonated GitHub repositories included fake tools branded as Claude, ComfyUI, AI coding assistants, Python security guides, and Rust frameworks.

Netskope discovered that the malicious payload arrives in a ZIP archive containing four files: two binaries (lua51.dll and compiler.exe), a batch script (Application.bat), and a text file (gc.txt). Despite its name, compiler.exe is a renamed LuaJIT interpreter that reads and executes the contents of the text file.

SmartLoader, a malware loader, is part of this campaign. It starts by sending a GET request to ip-api.com to collect the victim's IP address, country, city, time zone, and internet provider. The command-and-control IP is resolved from the Polygon blockchain at runtime, making it difficult for defenders to track.

The attackers use EtherHiding to query a smart contract through public RPC providers to get the address to contact. This setup allows them to move infrastructure without touching the malware itself. If a server gets blocked, they update the value stored in the smart contract, and every infected machine picks up the new address on its own.