This cybersecurity roundup highlights recent developments, including AI-assisted ransomware execution and exploits in network vulnerabilities. Key incidents involve ransomware experiments with Claude Skills, North Korean hacking activities, and significant policy and vulnerabilities updates. #ClaudeSkills #LazarusGroup
Keypoints
- Researchers demonstrated using Anthropic’s Claude AI skills to execute ransomware in controlled environments.
- An Array Networks vulnerability affecting AG secure access gateways has been exploited since August 2025 in Japan.
- North Korea’s Lazarus group is suspected of stealing around $30 million from Upbit cryptocurrency exchange.
- Akamai patched a critical HTTP request smuggling vulnerability, CVE-2025-66373, that could lead to data theft or redirection.
- The Iranian group MuddyWater developed a new backdoor called MuddyViper, targeting Israel and Egypt with sophisticated attacks.