IT Security News Roundup: 2026-11-18 Evening

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

Summary

Several threads run through this window: AI is changing both sides of security, with attackers using autonomous agents to build exploits and run intrusions, and defenders racing to keep up. Credentials and trust keep failing as the weak point, whether through reused admin logins that let ransomware gut a VMware estate, cloud identities that spread across providers, or a Russia-linked group turning one cloud breach into several more. Regulators and the human side appear too, from a record privacy fine in Italy to DORA's operational-resilience demands and the psychology of scam victims. Together these stories point to a reader's lesson: the basics of identity, configuration and patching still decide outcomes, and AI is shrinking the time available to act on them.

  1. Help Net Security interviews Christian Schwalbe, CEO of Deutsche Telekom IoT, on securing devices that stay in the field for 15 to 20 years. He covers the EU Cyber Resilience Act's impact on manufacturers and detecting threats from communication patterns without inspecting payloads.
  2. The Hacker News weekly recap rounds up a run of recent incidents, including zero-days in N-able, Windows, Chrome and a fridge, plus a default password that opened doors it should not have and a trusted update that delivered malware.
  3. TechCrunch reports that Anduril founder Palmer Luckey said Americans should keep three days of food, water and medicine on hand, and did not deny being a prepper.
  4. Huntress extends its 24/7 managed response to incidents from Microsoft Defender for Identity, Defender for Office 365, Defender for Cloud Apps and Entra ID Protection. The new visibility already helped uncover a token-replay campaign affecting 97 accounts at 79 organizations.
  5. Panda Security argues that cybercriminals can easily produce usable AI voice clones of people whose voices are widely available online, which makes voice-based impersonation and fraud a practical threat.
  6. OffSec's piece on DORA argues that the EU financial-sector resilience regulation is now simply how EU finance operates two years in, and asks whether security teams have the skills to sustain it.
  7. Ransomware group BlackNevas wiped a company's VMware estate in ten days, destroying ESXi hosts, virtual machines, vCenter appliances and backups. Kaspersky's response team says it used only valid admin credentials and legitimate VMware tools, with no exploited vulnerability.
  8. AWS Security Blog describes how to detect an adversary moving from one cloud identity into a second provider through a federated trust, by correlating signals that normally sit in separate services using AWS Security Hub.
  9. Dark Reading reports that China-nexus actors breached 10 organizations in autonomous, AI-driven attacks against advanced manufacturing, defense, and civil and human rights groups, using the Cairn framework.
  10. The Register reports a Linux TCP flaw, NoisyTCP, that lets a remote attacker crash an entire host, even when the vulnerable service runs inside a container, and that has existed for about 11 years.
  11. Malwarebytes advises readers not to let social media turn recent outages and attacks into a panic, since speculation about a bug can quickly be mistaken for a breach.
  12. The Hacker News reports Google's finding that threat actors have moved beyond manual prompting to autonomous multi-agent frameworks such as ARTEX, CAIRN, CyberStrikeAI and PentAGI, using them to turn software patches into working exploits.
  13. Dark Reading reports that the Russia-linked group Cold Relic uses data stolen from one cloud victim to target others through convincing impersonation and social engineering.
  14. Malwarebytes warns that scammers exploit GTA 6 launch hype with fake free codes that can cost victims their Google account, money or cryptocurrency.
  15. The Hacker News reports Java4Shell (CVE-2026-114938), disclosed by Calif, which allows unauthenticated remote code execution on Jira, Jenkins, Keycloak and other widely used Java applications, affecting Java 17.0.6 and later on AVX-512 VBMI x86-64 chips.
  16. Rapid7 details five ONLYOFFICE Docs vulnerabilities, found during Zimbra research, that let a view-only user chain flaws into root command execution inside the DocumentServer container. ONLYOFFICE fixed them in Document Server 10.0 on October 23, 2026.
  17. Securelist analyzes AuditTeam's OpsAgent implant, including its Rust rewrite, and shows how the group presents ransomware as a "forced audit" and demands a consulting fee to delete stolen data.
  18. Europol, with partners, dismantled the infrastructure behind the Gootloader, CastleRAT and StealC malware, seizing 235 servers and 257 domains and securing 6.4 million victim datasets.
  19. A honeypot-based research post reports that attackers brute-forced RDP credentials and then ran a previously undetected malware the researchers named Trojan.sysscan, while deception technology rerouted them into honeypots.
  20. Google's GTIG says Russia-linked Cold Relic (formerly COLDRIVER and UNC4057) used a 2026 cloud compromise at a U.S. organization to launch linked intrusions, including a second U.S. organization breach.
  21. The Italian data protection authority fined Meta €805 million over Facebook's gallery suggestions feature, which was on by default and accessed users' images, videos and metadata without clear consent. Meta has appealed.
  22. TechCrunch reports that Palmer Luckey warned AI-powered cyberattacks are outpacing defenses, urging organizations that outdated technology leaves them exposed: "It's time to start running."
  23. The Register sponsored piece asks whether AI's "Black October" has changed how teams secure code, noting that AI agents make vulnerabilities easier to find and exploit, and calls for new ways to find and fix them.
  24. OpenSourceMalware reports React2STEAL, which uses GitHub Actions to exploit React2Shell and has stolen AWS credentials, AI keys and .env files from more than 1,200 hosts that remain in public repositories.
  25. Help Net Security reports from IRISSCON 2026 that cyberpsychologist Dr. Nicola Fox Hamilton argues blaming scam victims deepens the harm and teaches them to stay silent next time.

Summaries written with AI from the linked source articles.

25
articles summarized
16
sources

Sources in this roundup

The Hacker News
4 article(s)
CySecurity News – Latest Information Security and Hacking Incidents
2 article(s)
Help Net Security
2 article(s)
Malwarebytes
2 article(s)
Security News | TechCrunch
2 article(s)
darkreading
2 article(s)
www.theregister.com – Articles
2 article(s)
AWS Security Blog
1 article(s)
Blog
1 article(s)
Huntress Blog
1 article(s)
News
1 article(s)
OffSec RSS Feed
1 article(s)
OpenSourceMalware Blog
1 article(s)
Panda Security Mediacenter
1 article(s)
Rapid7 Cybersecurity Blog
1 article(s)
Securelist
1 article(s)

Most-mentioned keywords

days
6 mention(s)
breach
3 mention(s)
years
3 mention(s)
access
2 mention(s)
cloud
2 mention(s)
cold
2 mention(s)
google
2 mention(s)
hackers
2 mention(s)

Sources

  1. Deutsche Telekom IoT’s CEO on securing devices that run for 20 years
  2. ⚡ Weekly Recap: N-able 0-Days, Windows 0-Days, Chrome 0-Days, Fridge 0-Days, and More Bad Days
  3. Palmer Luckey doesn’t deny that he’s a prepper
  4. Huntress Delivers 24/7 Response for Microsoft Defender Incidents Across Endpoints and Microsoft 365
  5. How easy is it for cybercriminals to create a realistic AI voice clone?
  6. DORA, Two Years In: Why Operational Resilience Is a Skills Question
  7. Ransomware Group Wiped an Entire VMware Estate in Ten Days Without Exploiting a Single Vulnerability
  8. Detecting cross-cloud lateral movement with AWS Security Hub
  9. China-Nexus Actors Breach 10 Organizations in AI-Driven Attacks
  10. NoisyTCP bug gives Linux servers a remote crash problem 11 years in the making
  11. Outage or attack? Don’t let social media fill in the gaps
  12. Google Finds Hackers Using AI Agents to Build and Run Exploits From Software Patches
  13. Cold Relic Turns One Cloud Breach Into Another
  14. GTA 6 scams could cost you your account, money, or crypto
  15. Java4Shell Flaw (CVE-2026-114938) Enables Pre-Auth RCE in Jira, Jenkins, Keycloak and More
  16. From view-only access to container root: five vulnerabilities in ONLYOFFICE Docs (FIXED)
  17. The forced audit: how AuditTeam’s OpsAgent framework masks ransomware as consulting
  18. 6.4 million victim datasets secured in international operation against malware networks
  19. The Oracle of Delphi Will Steal Your Credentials
  20. Russian COLD RELIC Hackers Use Stolen Google Workspace Data to Breach Second U.S. Organization
  21. Meta Fined €805 Million in Italy Over Facebook’s Access to Android Photo Galleries Without Consent
  22. Palmer Luckey warns AI-powered cyberattacks are outpacing defenses: ‘It’s time to start running’
  23. Has AI’s ‘Black October’ changed the way you secure your code?
  24. React2STEAL Uses GitHub Actions to Exploit React2Shell
  25. Once a scam ends the victim blaming begins