CISA Flags Actively Exploited Flaws in Wing FTP & Fortinet; Poland Thwarts Nuclear Site Attack

A new Armis report reinforces warnings about AI's role in cyberwarfare, stating it has reached a 'boiling point.' The study reveals 79% of IT leaders are concerned about nation-state AI attacks. Critically, 54% of organizations have already been hit by AI-led attacks, with 50% unable to fully secure their environments afterward. This highlights a dangerous disconnect and the challenge of 'machine speed' attacks, where adversaries operate faster than human defenders. The report also notes 52% of organizations face ransomware payouts exceeding their annual cybersecurity budget, indicating increased financial strain.

Security firm VulnCheck reports that a high-severity vulnerability, CVE-2026-20133 (insufficient file system access restrictions), is being overlooked due to misattribution of a PoC exploit. While a PoC was widely believed to target the previously reported zero-day CVE-2026-20127, it actually exploits CVE-2026-20133 and other flaws. This creates a dangerous blind spot, as organizations may believe they are secure after patching only CVE-2026-20127, leaving them vulnerable to CVE-2026-20133, which is also targeted by UAT-8616. Defenders must apply all relevant patches and verify detection rules for both CVEs.

Gartner forecasts that by 2028, half of all enterprise cybersecurity incident response efforts will be dedicated to incidents involving custom-built AI applications. This is due to their inherent complexity, dynamic nature, and deployment without adequate security testing, creating a new class of incidents for which most security teams are unprepared. This prediction significantly escalates the previously identified risk of insecure AI adoption, warning of substantial strain on resources, a skills gap, and the need for specialized AI security platforms.