Netcrook
#UAC-0099
When Malware Starts Speaking LLM: The New Evasion Trick Security Teams Must Fear
A reported campaign tied to UAC-0099 shows how adversarial text inside a script can pressure AI-assisted scanners into refusing analysis, turning model safety features into a defensive blind spot.
When a Script Comment Becomes a Trap for AI Security Tools
A VBScript sample tied to GuardBreaker shows how untrusted text inside code comments can be used to confuse AI-assisted malware analysis without changing how the script runs.
When Malware Talks Back: How a Prompt Can Jam AI Security Triage
A reported technique called GuardBreaker shows how malicious code may be written to confuse LLM-based analysis, turning AI guardrails into a potential delay mechanism rather than a defense.
The Fake Plugin Trick Turning a Text Editor into a Malware Disguise
A malicious add-on dressed as a Notepad++ plugin shows how attackers can weaponize trusted extension workflows on Windows without needing a flashy exploit.
When a Text Editor Turns Into a Loader: The Notepad++ Plugin Trap
A Windows campaign tied to UAC-0099 shows how a familiar plugin model can become an execution path for a malicious DLL.
Trusted Editor, Hidden Payload: Why a Notepad++ Wrapper Matters to Windows Defenders
A legitimate Notepad++ 8.8.3 executable has been linked to a reported malware chain, showing how trusted software can be used as a delivery wrapper without any claim that the editor itself is broken.