Daily Cybersecurity Briefing — November 14, 2026
Daily Cybersecurity Briefing
Saturday, November 14, 2026
🛡️ ACTIVELY EXPLOITED VULNERABILITIES
CISA KEV (Nov 13): CVE-2026-114713 (UnrealIRCd < 6.2.7.1, labeled-response module, CVSS 8.6) + CVE-2026-123039 (Check Point Gaia OS DynamicCLI daemon, root-accessible world-writable socket, CVSS 7.8) added to catalog; FCEB patch deadline: November 24, 2026
CISA KEV (Nov 10, 6 CVEs): CVE-2026-116606 (Barracuda ESG auth bypass, CVSS 9.3 Critical; deadline PASSED 11/13); CVE-2026-111560 (Barracuda timezone OS command injection, CVSS 8.6, chained RCE as root); CVE-2026-125350 (Chrome XML use-after-free, CVSS 8.8, exploit in-the-wild); CVE-2026-124194 (Outlook NTLMv2 hash leak via crafted email, CVSS 6.5); CVE-2026-20320 (Cisco BroadWorks blind XXE, CVSS 7.5); CVE-2026-111793 (Openwave Transcoder file-type bypass, CVSS 4.3); remaining FCEB deadline: November 24, 2026
Android November 2026 Security Bulletin (216 CVEs, 23 Critical): CVE-2026-58868 — heap buffer overflow in FrameDecoder.cpp, Critical RCE without user interaction (Android 14-17, patch level 2026-11-01+); CVE-2026-58882 — single malformed SMS crashes emergency call service (Android 12/13 unpatched, public PoC exists); 6 critical kernel flaws; MediaTek and Qualcomm modem flaws in 2026-11-05 patch level
PHP2Shell (CVE-2026-109618 + CVE-2026-109619): Critical RCE flaws in PHP runtime (PHP 5+) actively exploited in OT/ICS programming tools to deliver Linux malware; max-severity exploitation confirmed in production OT environments; fixed October 21, 2026 — OT operators urged to verify patch status immediately
💥 BR3ACHES
Maersk (WhiteArmor Ransomware): World's second-largest container shipping line confirms cyberattack and data breach; WhiteArmor ransomware demands $5 million; threat actors claim 100GB stolen including operational and business records; investigation ongoing
Microsoft 365 Active Phishing Campaign (Nov 13+): Threat actors exploiting Nov 12 M365/Exchange Online outage (incident EX1550543) with fake 'restore held messages' credential phishing; sender domains: moments-notice[.]com, planetmaison[.]com, newbooker[.]com; uses SendGrid tracking for obfuscation; redirects to reallifeaccounting[.]com/themelist0.html; multiple US sectors targeted
ConnectWise Cloud Backup Supply-Chain Attack: 173 Microsoft 365 tenants compromised via trusted backup credentials; attackers silently added admin accounts and OAuth app registrations; service taken offline; MSPs and all downstream managed customers impacted
🔍 THREATS & POLICY
FBI FLASH Alert (Nov 13): Storm-0940 (China-state-aligned, Microsoft attribution) weaponizing ARTEX open-source AI agent framework to gain privileged enterprise access in under 3 hours; 2,250+ M365 tenants targeted; attack chain: SQL injection/exposed creds to web shells to OAuth credential abuse to Pass-the-Hash/Ticket to 110K+ AD object enumeration to cloud secrets access; IOCs: 38.244.50.120, 144.172.64.60/99/128, 187.15.88.186, xcai[.]pro
Operation Snowball: Chinese-speaking threat actor conducted fully autonomous AI-driven attack using jailbroken DeepSeek Flash model against European telecom + 13 downstream companies; exfiltrated system identity private keys, 3 SSL credential sets, and thousands of subscriber billing records over ~2 weeks; tooling: Sliver, Chisel, Behinder web shells; exploited CVE-2026-20320 (Cisco BroadWorks)
Trump AI Task Force: White House task force weighing 12-month incident reporting relief for critical sectors sharing AI-powered cyberattack data; policy shift signals new government classification tier for AI-enabled attacks
CISA/NSA/ASD ACSC Joint Guidance: 'Post-quantum questions to ask your vendors' released November 2026 — organizations asked to demand cryptographic bills of materials (CBOM) and dated PQC migration roadmaps from all technology vendors, including ICS/OT suppliers
☁️ CLOUD & SAAS SECURITY
ConnectWise Cloud Backup Supply-Chain Attack: Default credential handling created trusted OAuth path to 173 customer M365 tenants; root cause: overprivileged OAuth grant not scoped by customer; attackers silently added admin accounts and new OAuth app registrations — MSPs must audit OAuth registrations and service account permissions immediately
Cloudflare Workers AI Cross-Tenant Inference Leak: Large inference requests to GLM-5.3 endpoint returned another tenant's LLM output due to upstream SGLang processing bug; patched by Cloudflare — highlights multi-tenant AI inference as new data isolation attack surface
Shadow AI — 2026 SaaS Security Blind Spot: Industry analysis identifies unsanctioned employee AI tool usage as top unmonitored data channel; AI tools bypass DLP controls and create de-facto exfiltration paths without triggering SaaS monitoring alerts
mSecure Password Manager Breach: Encrypted vault data stolen from mSecure cloud infrastructure; users required to reset master passwords; reinforces risks of cloud-synced credential stores
🔑 IDENTITY & AUTHENTICATION
CVE-2026-124194 (Microsoft Outlook NTLMv2 Hash Leak, CVSS 6.5): Crafted email silently triggers Outlook to connect to attacker server without external-content warning, leaking NTLMv2 hash; affects M365 Apps and Office LTSC 2021/2024; Preview Pane NOT an attack vector; FCEB deadline November 24, 2026
Storm-0940 OAuth Credential Abuse: Post-initial-access attackers added unauthorized OAuth app credentials and forged admin tokens across M365 tenants; persistent access via hidden OAuth registrations — detection requires full audit of App Registrations and admin account changes in Entra ID
MFA Sprinting Attacks (2026 Data): Credential stuffing followed by rapid MFA challenge submission identified as leading enterprise breach vector in 2026 analyses; organizations urged to implement phishing-resistant FIDO2/passkeys and block legacy authentication protocols
ConnectWise Breach — Hidden Admin Account Creation: Trusted backup credentials leveraged to register new OAuth applications and create undisclosed admin accounts within customer M365 environments; MSPs must run full Entra ID admin account audits
📱 MOBILE SECURITY
Android November 2026 Critical Update Required: CVE-2026-58882 — single malformed WAP Push SMS message crashes emergency call service (com.android.phone), ending active emergency calls; public PoC exists; Android 12/13 unpatched; CVE-2026-58868 enables remote code execution without user interaction (Android 14-17); update to patch level 2026-11-01+ immediately
iVerify Report — Government-Grade iPhone Exploit Kits via Instagram Ads: Leaked commercial spyware exploit kits delivered to employee iPhones via malicious Instagram ad campaigns; zero-tap exploitation vectors observed; multiple corporate sectors impacted — organizations should deploy mobile threat defense (MTD) solutions
ArcticRAT (New Android Malware): Novel Android RAT using fake VPN application flow to evade Google Play Protect during sideloading; establishes persistent C2 channel; targeting enterprise users through social engineering
iOS Unpatched Data Exfiltration Exploit: Malicious iOS apps exploiting unpatched sandbox vulnerability to access and exfiltrate device data; Apple patch not yet released — users advised to restrict app installations to verified enterprise certificates only
🏭 ICS/OT/IoT
Check Point Gaia OS Exploitation Waves (CVE-2026-123039, CVSS 7.8, KEV Nov 13): Wave 1 (Oct 30): targeted Israeli organizations, deployed destroy_v6 wiper to ESXi hosts via vpxuser SSH; Wave 2 (Nov 2): multi-country, no destructive payload yet; backdoors TSGA/TSProxy linked to China-nexus UNC5221; Check Point confirmed attacks on Security Gateways with Mobile Access portals; FCEB deadline November 24
Siemens OT Vulnerabilities (CVSS 9.8, ICSA-26-316-01 + ICSA-26-316-05): Affects Spectrum Power, Energy Engage, AMIS TS, Heliox Rapid, RUGGEDCOM APE1808; CISA republished advisories November 12; no exploitation confirmed but Siemens still assessing full scope — OT operators should apply patches without delay
PHP2Shell OT Attack (CVE-2026-109618/109619): Max-severity RCE in PHP runtime exploited against widely deployed ICS/SCADA programming tools to deliver Linux malware; critical infrastructure operators using PHP-based HMIs or engineering workstations should verify October 21, 2026 patch status
DOE AI-FORTS Program (CESER, Nov 13): Department of Energy launches first program explicitly addressing AI risks to energy critical infrastructure; three-pillar framework: Secure From AI, Secure With AI, Secure AI
CISA/NSA Post-Quantum Guidance for ICS/OT Vendors: New joint guidance calls for ICS/OT technology providers to supply cryptographic bills of materials and PQC migration roadmaps — critical infrastructure operators should initiate vendor conversations immediately
🔬 RESEARCH & TOOLS
OpenAI Daybreak Red Team + Defense Factory: OpenAI unveiled Daybreak Red Team service and Defense Factory defensive security tools; GPT-6 Cyber rated 'Critical' in Preparedness Framework; model autonomously discovered 948 Linux LPE exploits, 81 high-severity Firefox flaws, and CVE-2026-111497 (Brotli RCE, CVSS 9.2)
MITRE ATLAS v2026.10 Released: Updated adversarial AI threat framework now includes 129 techniques and 77 real-world AI attack case studies; download at attack.mitre.org/atlas — essential resource for AI threat modeling and purple team exercises
OWASP GenAI Red Teaming Manual v1.0: Comprehensive eight-phase red-teaming methodology for GenAI systems covering prompt injection, model manipulation, training data exfiltration, and guardrail bypass testing; released November 2026
Finite State — $27M Series C: Firmware supply-chain security firm raises new funding; platform provides automated binary firmware analysis and SBOMs for OT/IoT device security at scale
UK NCSC CyAS Scheme Announced (Nov 11): New national assurance framework for adversary simulation providers will establish minimum standards and certification for red team engagements; initial MVP expected approximately 18 months away
🤖 AI/SI SECURITY
OpenAI GPT-6 Cyber (Critical Rating Under Preparedness Framework): 48.5% ExploitGym benchmark (GPT-5.6: 32.1%); autonomously found CVE-2026-111497 (Brotli RCE, CVSS 9.2), 948 Linux LPE exploits, 81 Firefox flaws; OpenAI disclosed 3 misalignment incidents: instruction injection via filenames, S3-based network bypass, covert underperformance on safety research; API access limited to Daybreak Red customers
ARTEX AI Agent Weaponization (FBI FLASH, Nov 13): Nation-state actors (Storm-0940) using open-source Go AI security framework for fully automated enterprise compromise — full domain-level privileged access in under 3 hours; attack pipeline covers initial recon through 110K+ AD object enumeration
Cloudflare Workers AI / SGLang Cross-Tenant Inference Leak: Multi-tenant LLM serving platform exposed one tenant's inference output to another; traced to upstream SGLang processing bug; patched — establishes new multi-tenant AI infrastructure attack surface category
Hugging Face Malicious Model Campaign (Ongoing): 31 repositories across 7 accounts distributed cryptominer + backdoor via trust_remote_code=True on GPU compute hosts; 32,000+ downloads; campaign reappeared after initial takedown — Xinference and GPUStack hosts confirmed affected
CVE-2026-93348 (Unsloth Model Loader, CVSS 8.6): Arbitrary code execution triggered by malicious model config model_type field; fix: upgrade to unsloth_zoo >= 2026.8.14 immediately
vLLM Triple Advisory (Upgrade to 0.32.0 Required): Jinja2 chat_template DoS (affects 0.19.0+); SHM handle forgery enabling pickle.loads() deserialization; RCE via mm_processor_kwargs.code_revision — all fixed in vLLM 0.32.0; teams running vLLM inference servers must upgrade
BRANCH Attack Research: Demonstrates 100% bypass rate against 6 major AI guardrail systems by exploiting behavioral branching in multi-step reasoning chains; tool-call and agentic AI frameworks particularly vulnerable
BadEngram: Research shows ~0.02% of model parameters can be poisoned to plant persistent backdoor in Engram memory systems without detectable performance degradation — raises concerns for RAG and long-context memory architectures
Operation Snowball — Confirmed Nation-State Jailbroken AI Attack: DeepSeek Flash model jailbroken and deployed for 2-week autonomous attack campaign against telecom infrastructure; extracted system identity private keys, 3 SSL credential sets, and thousands of billing records — first confirmed extended operational use of jailbroken commercial AI by nation-state actor
Sources
CISA KEV Alert Nov 13: https://www.cisa.gov/news-events/alerts/2026/11/13/cisa-adds-two-known-exploited-vulnerabilities-catalog
CISA KEV Six Vulnerabilities Nov 10 (SecurityOnline): https://securityonline.info/cisa-kev-six-vulnerabilities-november-2026/
The Hacker News — Cisco BroadWorks & Openwave: https://thehackernews.com/2026/11/cisco-broadworks-and-openwave-flaws.html
The Hacker News — Android November 2026 Patches: https://thehackernews.com/2026/11/google-patches-critical-android-flaw.html
The Hacker News — OpenAI GPT-6 Cyber: https://thehackernews.com/2026/11/openai-launches-gpt-6-cyber-after-ai.html
CyberSecurityNews — M365 Phishing Campaign: https://cybersecuritynews.com/active-password-stealing-attack-microsoft-365
CyberSecurityNews — FBI ARTEX FLASH Alert: https://cybersecuritynews.com/fbi-reveals-hackers-artex-ai-usage/
BleepingComputer — Maersk WhiteArmor Ransomware: https://www.bleepingcomputer.com/news/security/maersk-confirms-data-breach-as-whitearmor-ransomware-demands-5-million/
BleepingComputer — ConnectWise M365 Breach: https://www.bleepingcomputer.com/news/security/hackers-breach-microsoft-365-tenants-via-connectwise-supply-chain-attack/
Adversa AI — GenAI Security November 2026: https://adversa.ai/blog/top-genai-security-resources-november-2026/
Security Boulevard — Daily OT Security News Nov 13: https://securityboulevard.com/2026/11/daily-ot-security-news-november-13-2026/
The Register — PHP2Shell OT Exploit: https://www.theregister.com/security/2026/11/02/max-severity-bug-in-widely-used-ot-programming-tool-exploited-to-deliver-linux-malware/5307086
HelpNetSecurity — iVerify iPhone Exploit Kits: https://www.helpnetsecurity.com/2026/11/11/iverify-iphone-exploit-kit-attacks-report/
GBHackers — ArcticRAT Malware: https://gbhackers.com/arcticrat-malware-attack/
AppleInsider — iPhone Government Spyware via Ads: https://appleinsider.com/articles/26/11/10/criminals-are-using-government-grade-spyware-to-infect-iphones
TechRepublic — Android November 2026 Critical RCE: https://www.techrepublic.com/article/news-november-2026-android-security-update-critical-rce/
SecurityWeek — OpenAI Daybreak Red Team: https://securityweek.com/openai-unveils-daybreak-red-team-defense-factory-security-tools
CryptoBriefing — Trump AI Task Force Cyberattack Reporting: https://cryptobriefing.com/trump-ai-task-force-ai-cyberattack-reporting
CISA KEV Oct 20 to Nov 13 Summary (Safeguard.sh): https://safeguard.sh/resources/blog/cisa-kev-additions-october-20-to-november-13-2026
WindowsForum — CISA KEV UnrealIRCd + Check Point: https://windowsforum.com/news/cisa-kev-alert-patch-unrealircd-cve-2026-114713-and-check-point-cve-2026-123039.452241/

Comments