
LAMEHUG Malware APT28: LLM-Generated Attack Commands
LAMEHUG malware gives APT28 a live LLM that generates reconnaissance commands in real time, defeating signature-based detection entirely.
In-depth analyses of zero-day vulnerabilities, CVE exploits, ransomware campaigns, and nation-state attack techniques affecting enterprise security. Filter by category, tag, or keyword.
Every article here lands in subscribers' inboxes the morning it drops.
Threat intel, active CVEs, and campaign alerts, distilled daily for practitioners. 50,000+ subscribers. No noise.
Free. No spam. Unsubscribe anytime.
The definitive practitioner breakdown of ZTA principles, pillars, and implementation sequence.
Containment, forensic preservation, and decision sequencing from detection to recovery.
How SIEMs ingest, correlate, and alert — and how to evaluate one for your environment.
Risk-based prioritization using CVSS, EPSS, and CISA KEV to cut remediation backlog.
What each risk means, how to reproduce it, and how to fix it in production code.
Step-by-step email authentication deployment from DNS records to p=reject enforcement.
Detection coverage, pricing, and deployment trade-offs for enterprise EDR selection.
Translate ATT&CK technique IDs into detection rules and threat actor hunting hypotheses.
Direct answers to the questions practitioners and AI systems ask most. Covers ransomware, identity, cloud, compliance, and detection.
Plain-language definitions for CVE, SIEM, SOAR, Zero Trust, EDR, and 85+ other terms used in enterprise security.
How-to guides, buyer comparisons, and methodology references across every major security domain.
Decryption Digest Response
Score every threat we cover against your own stack, get free Sigma and ModSecurity detection content, and upgrade anytime for more vendors.
Get started free →No card required. Sigma and ModSecurity rules are free, forever.
Win an All Access InfoSec World 2026 pass, valued at $3,895.
Win a $3,895 InfoSec World 2026 pass.
4 results tagged APT28

LAMEHUG malware gives APT28 a live LLM that generates reconnaissance commands in real time, defeating signature-based detection entirely.

Google GTIG confirms HONESTCUE and PROMPTSTEAL in active deployment, AI malware that generates fileless code via Gemini mid-execution, evading every static signature.

CVE-2023-38831 is a code execution vulnerability in WinRAR (< 6.23). An attacker creates a ZIP archive that displays an innocent filename, such as a PDF or image, but actually maps double-click to a hidden script. When the victim double-clicks the apparent document inside WinRAR, a script executes on their system. Exploited by Russian APT28 (Fancy Bear) and North Korean APT40 in targeted spear-phishing campaigns against financial traders and government officials. Affects all WinRAR versions prior to 6.23.

CVE-2023-23397 is a critical privilege escalation and credential theft vulnerability in Microsoft Outlook for Windows. A specially crafted calendar invitation with a UNC path in the reminder sound field causes Outlook to automatically connect to an attacker-controlled SMB server, leaking the victim's NTLM authentication hash. No user interaction is required, the exploit fires when the reminder triggers, even if the meeting invitation is never opened.