⚡ Weekly Recap: Linux Kernel Flaws, AI Malware Tricks, Turla Backdoor, Infostealers and More

This week was a reminder that attackers do not always need big tricks. One small mistake, one old access path, one missed patch, and suddenly the door is open. The noise is not all noise, either. Forums are talking, researchers are finding easy cracks, and defenders have more cleanup waiting. Here’s the full Monday recap. ⚡ Threat of the Week New DirtyClone Linux Kernel Flaw Lets Local

The recap highlights new AI-linked threats, notably Gaslight macOS malware and a Rust-based macOS implant that embed prompt injection payloads specifically to mislead AI-assisted malware analysis tools into aborting or refusing analysis.[2][4] It also reports serious indirect prompt injection risks in agentic IDEs and coding agents, where attacker-controlled but seemingly benign repositories can trigger tool access, code execution, file operations, and network calls.[2][4] From a CyberSE.AI perspective, these demonstrate that AI-powered security and coding tools can be turned into attack surfaces: organizations should treat AI agents as high-privilege components, enforce strict tool- and repo-access controls, and continuously red-team agent workflows to identify and mitigate indirect prompt injection paths before they lead to compromise.