IT Security News Roundup: 2026-11-19 Morning

IT Security News: Morning roundup, 2026-11-19

Summary

This window was dominated by vulnerabilities in widely deployed software and by AI cutting both ways: attackers used an AI tool against the European Commission, while researchers used AI to find nginx flaws and to formally verify the Signal protocol. Several stories also turn on trust and credentials, from ransomware gangs posing as IT support to a supply-chain bypass of GitHub SHA pinning and a leaked sticky note. Together they show that the same weak points, such as exposed services, stolen logins and poorly vetted tooling, are being exploited faster than many organizations can patch or verify them.

  1. Five new suspects, four men and one woman, were added to the EU Most Wanted list over the 2022 TEFAF art fair robbery in Maastricht, according to Dutch prosecutors and police. The announcement shows the cross-border investigation is still active years later.
  2. Google released Gemini 4 Neon, a coding model for multi-file debugging and refactoring that it prices at $3 per million input tokens and $15 per million output tokens. Google says it costs about 32% less per task than Gemini 4 Carbon, which could push enterprises toward cheaper coding agents.
  3. Ukrainian police, working with the FBI over more than two years, arrested four key members of a ransomware group that took over $40.5 million from at least 28 companies. The group broke in by phoning victims while posing as IT support staff.
  4. A sticky note with login details on a monitor exposed credentials that gave access to a company's systems. The incident is a reminder that physical security lapses can undermine strong digital controls.
  5. Oregon is seeking at least $2.035 billion from OpenAI, alleging that ChatGPT failed to protect teenagers. The case could set a precedent for how child-safety obligations apply to AI chatbots.
  6. Italy fined Meta €805 million over aggressive access to users' Facebook photos, and the privacy violation is still ongoing for some users. The case highlights how app permissions can quietly extend far beyond what users expect.
  7. CISA added four vulnerabilities to its Known Exploited Vulnerabilities catalog, affecting Synology DiskStation Manager, BerriAI LiteLLM, Apache Axis and Vanderbilt REDCap. Evidence of active exploitation links them to ransomware, cryptomining and LLMjacking campaigns, so administrators should prioritize patching them.
  8. Xint Research disclosed nginx2root, five nginx vulnerabilities and one Linux kernel bug, each with a working exploit. The most severe flaw, CVE-2026-123007, lets an unauthenticated attacker run code, and the research was produced with AI assistance.
  9. Google DeepMind used agentic AI to formally verify several key cryptographic guarantees of the Signal protocol in a four-week proof-of-concept project. The result independently supports the security of messaging apps such as Signal, Google Messages and WhatsApp.
  10. A cyberattack on the European Commission disrupted staff services for more than 10 days, and CERT-EU found that the attackers used an AI tool called ARTEX. Most affected applications are back online, but the incident shows how AI tools can be turned into attack instruments against major institutions.
  11. Apache disclosed CVE-2026-129448, an actively exploited remote code execution flaw in Apache Roller that needs no login. It affects versions 4.0.0 through 6.1.4 and has a CVSS score of 9.8, so administrators running affected versions should upgrade without delay.
  12. Researchers at the University of Osaka observed an unusual quantum state at the boundary between a one-atom-thick material and copper, where electrons behave as if they were heavy. The finding is basic physics research with no direct security impact.
  13. F5 released patches for five critical NGINX2Root flaws that could let remote, unauthenticated attackers execute arbitrary code. Public proof-of-concept exploits are now available, which raises the urgency for NGINX operators to update.
  14. A suspected Chinese advanced persistent threat group rebuilt its EggStreme malware and supporting infrastructure after Bitdefender exposed them. The quick rebuild shows that publicizing a campaign alone does not end it.
  15. An international operation during a hackathon in Amsterdam, supported by Europol and Eurojust, identified 51 potential human trafficking victims and five suspects. The results show how coordinated online investigations can locate people at risk.
  16. Researchers detailed a now-patched Apple flaw that let apps on iPhones and iPads record audio even after users denied microphone permission. Fixes shipped on August 17, and neither Apple nor the researchers reported that it was used in attacks.
  17. Palo Alto Networks agreed to acquire Applied Compute, a platform for post-training and running open-weight AI models, in an all-stock deal valued at $4.5 billion. The move signals a push to secure enterprise AI adoption across both open and closed models.
  18. Attackers found a way around GitHub SHA pinning, a common defense against supply-chain attacks, by hijacking the Trivy scanner in March 2026. The malicious code was disguised as GitHub's official checkout tool, so pinned versions alone are no longer sufficient protection.
  19. Scientists developed a method to measure uncertainty in quantum simulations and provide verifiable error margins, demonstrated on a 51-qubit system. The work addresses a growing problem as quantum machines outgrow conventional tools for checking their results.
  20. Kaspersky reported Operation ShadowCloud, a campaign against government sectors in Asia and Africa that uses previously undocumented GitDoor and GraphDoor backdoors. The malware abuses Microsoft SharePoint and GitHub as covert command-and-control channels, making its traffic harder to block.
  21. Scammers are exploiting interest in GTA 6 with fake giveaways, game keys that may never arrive and sign-in or wallet-connection lures. Malwarebytes researchers say the schemes aim to steal Google passwords, money and cryptocurrency.
  22. N-able disclosed that unknown attackers gained unauthorized access to a server in its Mail Assure and Spam Experts hosted cloud, disabling management interfaces and APIs in three regions. The investigation is ongoing, and customers depending on that email security service were affected.
  23. Cognizant announced it will use the full Qualys portfolio in a managed security service built for AI-speed threat detection, prioritization and remediation. The offering targets enterprises that need to fix newly disclosed zero-day vulnerabilities quickly.
  24. Market research firm Ipsos disclosed a cybersecurity incident and brought in external specialists to determine its cause and scope. The company has not yet publicly detailed what data, if any, was affected.
  25. NVIDIA said it is investigating claims of a breach after the Gentlemen ransomware group listed the company on its dark web data leak site. The listing means the group is pressuring NVIDIA publicly, whatever the outcome of the investigation.

Summaries written with AI from the linked source articles.

25
articles summarized
16
sources

Sources in this roundup

The Hacker News
5 article(s)
Help Net Security
3 article(s)
Hacking News — ScienceDaily
2 article(s)
News
2 article(s)
www.theregister.com – Articles
2 article(s)
BleepingComputer
1 article(s)
Cyber Security News
1 article(s)
GBHackers Security | #1 Globally Trusted Cyber Security News Platform
1 article(s)
InfoWorld
1 article(s)
Malwarebytes
1 article(s)
Qualys Security Blog
1 article(s)
Schneier on Security
1 article(s)
Security Affairs
1 article(s)
Security Archives – TechRepublic
1 article(s)
securityweek
1 article(s)
www.infosecurity-magazine.com
1 article(s)

Most-mentioned keywords

flaws
3 mention(s)
hackers
3 mention(s)
nginx
3 mention(s)
ransomware
3 mention(s)
root
3 mention(s)
breach
2 mention(s)
critical
2 mention(s)
five
2 mention(s)

Sources

  1. Five TEFAF robbery suspects added to EU Most Wanted
  2. Google launches Gemini 4 Neon, touting lower costs for complex AI coding tasks
  3. Ukraine arrests key members of ransomware gang that extorted $40.5 million posing as IT support
  4. Sticky note on monitor provided keys to the company’s kingdom
  5. Oregon Sues OpenAI for $2.035B Over ChatGPT Teen Safety
  6. Facebook may have access to your entire photo gallery
  7. CISA Warns of 4 Exploited Flaws Linked to Ransomware, Cryptomining, and LLMjacking Attacks
  8. nginx2root:AI Discovers Five Critical nginx Flaws and a Path to Root
  9. Formal Verification of the Signal Protocol Using AI
  10. Hackers Target European Commission with ARTEX AI Tool in Major Cyberattack
  11. Apache Roller Vulnerability Allows Remote Code Execution Through Unsafe XML-RPC Deserialization
  12. A material just one atom thick reveals a strange quantum secret
  13. F5 Patches Critical NGINX2Root Flaws as PoC Exploits Go Public
  14. Suspected Chinese APT Rebuilds EggStreme Malware After Exposure
  15. 5th EMPACT Hackathon leads to the detection of 51 potential victims
  16. Researchers Detail iPhone Flaw That Let Apps Record Audio Despite Denied Microphone Permission
  17. Palo Alto Networks to acquire Applied Compute for $4.5 billion
  18. Hackers Found a Way Around GitHub SHA Pinning. This Webinar Explains How to Close the Gap
  19. Quantum computers are getting too powerful to check. Scientists have a solution
  20. New GitDoor and GraphDoor Backdoors Turn GitHub and SharePoint Into Covert C2 Channels
  21. GTA 6 scams are after your Google password, money, and crypto wallet
  22. Hackers Breach N-able Cloud Server, Disrupting Email Security Management Across 3 Regions
  23. Cognizant and Qualys Redefine AI-Era Managed Security Service
  24. Opinion pollster Ipsos 'fesses up to 'cybersecurity incident'
  25. NVIDIA investigating Gentlemen ransomware breach claims