IT Security News Roundup: 2026-11-18 Afternoon

IT Security News: Afternoon roundup, 2026-11-18

Summary

This window was dominated by AI showing up on both sides of security: attackers using an automated framework to gain privileged access, phishing that can reach AI coding agents, and a breach at an AI vendor that exposed driver data. Vulnerabilities in widely used components also ran through the period, including critical NGINX and Node.js-related flaws, an actively exploited ScreenConnect bug, and several Siemens advisories. Taken together, the stories point to a common concern: the tools organizations adopt quickly, whether AI services, remote access software, or web servers, often become the path in, so their access, configuration, and patching need the same scrutiny as everything else.

  1. Wiz has launched an AI Security Agent, currently in public preview, that analyzes application code and cloud permissions to surface hidden AI agents and attack paths. Wiz reports improved discovery, but that claim has not been independently verified.
  2. The article warns that anyone can publish a public AWS AMI, so names and descriptions cannot be trusted. Malicious or misconfigured images can ship vulnerable software, malware, exposed credentials, or insecure settings that compromise workloads at launch.
  3. Norway-based energy company Equinor is investigating a cyberattack that has affected its IT systems. It says production was not disrupted, but it has no immediate timetable for full recovery.
  4. Wilmington School District 209U in Illinois is investigating a possible ransomware attack and data breach. The superintendent told staff and families that experts are working to determine its extent and whether sensitive information was accessed.
  5. The FBI warns that ARTEX, a new framework, automates several stages of a cyberattack. It says hackers used the tool to gain privileged access in under three hours.
  6. Malware-Traffic-Analysis reports a SmartApeSG ClickFix campaign delivering the CNCmachineRMS remote access trojan. The available text is limited, so further details could not be confirmed from the source.
  7. Dorsett Controls InfoScan, used in water and wastewater systems, has six vulnerabilities in versions 1.45.0 and earlier. They could expose sensitive data or stored credentials, or let attackers manipulate network traffic. Version 1.46 contains the fix.
  8. Microsoft's lab research shows that device code phishing, a technique that applies to any device-flow client, can extend to AI coding agents. GitHub Copilot CLI is used as the illustrative example.
  9. This OffSec piece is promotional content about using Upskill Program results to plan next year's training budget. It offers guidance on workforce development rather than reporting a security incident.
  10. The Register's partner content says insurers are tightening underwriting and now expect evidence that backups can actually be restored. A backup job showing green is no longer enough.
  11. Two critical NGINX flaws, CVE-2026-123633 and CVE-2026-123007, could let remote attackers run code on vulnerable web servers without logging in. Both require HTTP/2 connections, and one affects certain reverse proxy configurations.
  12. This Darknet guide explains how to install ProjectDiscovery's httpx 1.12.0, distinguish it from Kali's httpx-toolkit and Python's httpx, and use it to probe live web servers from subfinder output. It is a tooling how-to rather than a news item.
  13. IT Security Guru cites Black Duck's 2027 AI Readiness Report, which analyzed 65.5 million application security findings. It argues that severity scores fail teams, widening the gap between what is found and what is fixed.
  14. Qualys's ROCky is now on the Google Cloud Marketplace for Gemini Enterprise, letting teams ask natural-language questions about vulnerabilities and assets, with answers ranked by real risk. The post is vendor promotion rather than a reported incident.
  15. Siemens has published an advisory covering many of its industrial software products affected by Node2Shell, a critical out-of-bounds write in the Node.js Brotli decoder tracked as CVE-2026-111497. Fixes are in preparation for some products, and mitigations are recommended where none exist yet.
  16. A critical ScreenConnect flaw is being exploited to run malicious files on technician accounts at managed service providers. Security researchers counted almost 690,000 file transfer requests in less than a month.
  17. Siemens ROS# before version 2.3.1 contains a code injection flaw, CVE-2026-109097, that could let an attacker inject arbitrary code through a crafted message definition file. The CVSS score is 8.2, and the fix is to update to 2.3.1.
  18. A ride-hailing company says stolen credentials at an AI vendor let attackers take license data belonging to Uber drivers. The breach is a reminder that third-party AI tools can expose sensitive personal records.
  19. Siemens Simcenter Femap has an out-of-bounds read flaw, CVE-2026-64629, triggered by specially crafted X_T files. It could crash the application or allow code execution, and the fix is version V2606.0002.
  20. Cybersecurity Dive reports that healthcare organizations are adopting AI quickly while security teams are stretched too thin to monitor it. Unmonitored shadow AI tools are linked to breaches.
  21. Siemens SIMATIC products, including WinCC and Process Historian, are affected by CVE-2026-58114, a weakness in the deprecated NTLM protocol. An on-path attacker could relay authentication and impersonate users, and some affected products have no fix planned.
  22. Malware is being distributed through websites offering a free PC version of GTA VI and is designed to steal passwords, browser data, and cryptocurrency information. The article notes the game's console launch is scheduled for November 19.
  23. CrowdStrike is extending its Agentic SOC to detect and stop insider threats from AI agents. The available text is limited to the headline and announcement, so product specifics could not be confirmed.
  24. Google reports that Russian state-backed hackers use information stolen from cloud accounts to impersonate trusted contacts. They then launch new espionage attacks against people in their victims' professional networks.
  25. Rapid7's Surface Command now integrates with Claude's Compliance API, bringing organization, user, role, group, and settings data into one view for access reviews. IBM's 2025 report found that 97% of organizations with an AI-related breach lacked proper AI access controls. The integration provides administrative visibility only and does not ingest conversation content.

Summaries written with AI from the linked source articles.

25
articles summarized
16
sources

Sources in this roundup

All CISA Advisories
5 article(s)
Cybersecurity Dive – Latest News
3 article(s)
DataBreaches.Net
2 article(s)
Qualys Security Blog
2 article(s)
www.theregister.com – Articles
2 article(s)
Blog
1 article(s)
Cyber Security News
1 article(s)
Darknet – Hacking Tools, Hacker News & Cyber Security
1 article(s)
Hackread – Cybersecurity News, Data Breaches, AI and More
1 article(s)
IT SECURITY GURU
1 article(s)
Malware-Traffic-Analysis.net – Blog Entries
1 article(s)
Microsoft Security Blog
1 article(s)
OffSec RSS Feed
1 article(s)
Rapid7 Cybersecurity Blog
1 article(s)
The Record from Recorded Future News
1 article(s)
esecurityplanet
1 article(s)

Most-mentioned keywords

siemens
4 mention(s)
access
3 mention(s)
data
3 mention(s)
risks
3 mention(s)
agent
2 mention(s)
cloud
2 mention(s)
code
2 mention(s)
cyber
2 mention(s)

Sources

  1. Wiz Launches AI Security Agent to Find Hidden Risks in Enterprise AI Apps
  2. Understanding the Risks of Public AWS AMIs: Hidden Dangers of Malicious and Misconfigured Images
  3. Energy giant Equinor investigates cyberattack impacting IT systems
  4. IL: Wilmington School District 209U Reports Ransomware Attack And Possible Data Breach
  5. FBI Says ARTEX AI Tool Helped Hackers Gain Privileged Access in Under 3 Hours
  6. 2026-11-12: SmartApeSG ClickFix campaign pushes CNCmachineRMS RAT
  7. Dorsett Controls InfoScan
  8. The risks of device code authentication for AI coding agents
  9. Use Upskill Program Results to Plan Next Year’s Training Investment
  10. Cyber insurers now want proof you can actually recover
  11. NGINX HTTP/2 Flaws Could Allow Memory Corruption and Remote Code Execution on Web Servers
  12. httpx – ProjectDiscovery’s HTTP Toolkit for Probing Live Hosts
  13. Fix what matters, not what’s easy: why severity scores are failing security teams
  14. Run Your Qualys Program by Conversation, Inside Google Cloud’s Gemini Enterprise
  15. Siemens Industrial Software Products
  16. Critical flaw in ScreenConnect exploited to run malicious files on MSP technician accounts
  17. Siemens ROS#
  18. Hackers smash 'n' grab AI vendor: Take off with license data of Uber drivers
  19. Siemens Simcenter Femap
  20. Healthcare firms fail to monitor AI tools as shadow AI leads to breaches
  21. Siemens SIMATIC
  22. Fake GTA 6 PC Downloads Spread Malware Targeting Crypto Wallets and Passwords
  23. CrowdStrike Extends the Agentic SOC to Detect and Stop Agent Insider Threats
  24. Russian cyber-espionage group used stolen cloud data to hack into organizations, Google says
  25. Who Has Access to Claude? Bring AI Identity and Access Context Into Surface Command