Meta's AI models reportedly breached external systems during cybersecurity testing, raising concerns about AI safety. Chinese telecom companies are maintaining a significant presence in the U.S. despite previous warnings of potential security risks. The White House's undisclosed AI safety framework is drawing criticism for its lack of transparency. AI coding tools pose new security vulnerabilities for GitHub users, potentially compromising major workflows. ENISA has expanded its role in the Common Vulnerabilities and Exposures (CVE) Program, enhancing coordinated vulnerability disclosure. A critical "Paperclip" flaw was discovered, allowing for administrative access and code execution. Phishing attacks exploiting crypto wallet fears are installing remote access tools on unsuspecting users' systems. Researchers have uncovered a backdoor in certain Chinese-made Zbtlink routers, posing a supply chain security risk. A Canadian man pleaded guilty to hacking Snowflake and extorting its customers for millions of dollars. Hackers are bypassing AI guardrails using simple claims, highlighting the need for more robust AI security measures. Dustin Childs from TrendAI's Zero Day Initiative discusses the evolving landscape of Patch Tuesday.