AI is Everywhere, But CISOs are Still Securing It with Yesterday's Skills and Tools, Study Finds

The collection summarizes a wide range of active threats and security developments: organizations remain underprepared to secure AI systems; North Korean Konni actors use spear‑phishing to deploy EndRAT and propagate via KakaoTalk; GlassWorm campaigns abuse stolen GitHub tokens and Open VSX/extension dependencies to inject malware into Python projects; ClickFix campaigns deliver MacSync macOS stealer; OpenClaw autonomous AI agent has weak defaults enabling prompt injection and data exfiltration; Google patched two Chrome zero‑days (including CVE-2026-3909) exploited in the wild; Veeam fixed多个高