Koffei

Daily CyberNews · 03/11/2026

‹ 02/11/202604/11/2026 ›
Today in 30 seconds
  • 🔴 Update Node-RED on SIMATIC IoT2050 to 4.3.4.1: CVE-2026-58115 is exploited.
  • 🔴 Remove the CryptoPayment Gateway WordPress plugin: CVE-2026-81648 is attacked and has no fix.
  • 🔴 Update Synology DSM: C3RB3R ransomware is hitting unpatched NAS devices.
  • 🔴 Check Danfoss refrigeration controllers for outdated software and default passwords after attacks in eight countries.
  • 🟠 Northflank customers: rotate every environment-variable secret, keys are being misused.

Vulnerabilities 6

🔴 Exploited or active

Siemens SIMATIC IoT2050 / Node-RED missing-authentication RCE

CVE-2026-58115Siemens SIMATIC IoT2050CVSS 10.0exploited

WhatMissing authentication in Node-RED on SIMATIC IoT2050 Advanced lets an unauthenticated attacker take full control; honeypots have seen exploitation since Oct 30.

DoUpdate Node-RED to 4.3.4.1 or later per Siemens advisory SSA-834709.

Details and sources
CVE-2026-58115 (CVSS 10.0, CWE-306 missing authentication) lets an unauthenticated remote attacker take full control of SIMATIC IoT2050 Advanced devices running Node-RED below v4.3.4.1 via the exposed HTTP interface. VulnCheck honeypots have observed active exploitation since Oct 30, 2026. Fixed in Node-RED 4.3.4.1+ (Siemens advisory SSA-834709). The Hacker News (source)
🔴 Exploited or active

CryptoPayment Gateway WordPress plugin — unauthenticated file deletion / wallet-credential exposure

CVE-2026-81648CryptoPayment Gateway WordPress pluginno fix

WhatUnauthenticated attackers can delete server files, overwrite payment settings and recover wallet credentials in cleartext; in-the-wild attempts were observed.

DoRemove the plugin (versions 1.2.1 to 1.2.2) outright, since no fix exists.

Details and sources
CVE-2026-81648 (CVSS 10.0, missing authorization) in versions 1.2.1–1.2.2 lets unauthenticated attackers delete arbitrary server files, overwrite payment configuration, and recover wallet credentials in cleartext. In-the-wild attempts observed by Defused Cyber. No fix is available — the plugin was pulled from the WordPress directory back on Aug 28, 2026, so affected sites must remove it outright. The Hacker News (source)
🔴 Exploited or active

Ivanti Connect Secure — three-flaw chain gives unauthenticated root (UTA0576)

UTA0576CVE-2026-104935CVE-2026-111357CVE-2026-118527

WhatUTA0576 chains an auth bypass, privilege escalation and file upload in Ivanti Connect Secure for unauthenticated root and credential theft; two flaws are in KEV.

DoUpgrade to Ivanti Connect Secure 25.1.1.3 or 22.8R2.6; federal deadlines Nov 2 and Nov 13.

Details and sources
A previously undocumented actor Volexity calls UTA0576 chains an auth-bypass flaw, CVE-2026-104935 (CVSS 8.6, spoofing-based auth bypass in ICS before 25.1.3.0), with a local privilege-escalation issue, CVE-2026-111357 (CVSS 7.0 per NVD / 8.8 per other reporting, DSCache trust-handling flaw affecting 22.8R2.4), and a predictable-identifier file-upload bug, CVE-2026-118527 (CVSS 4.3), to reach unauthenticated root code execution and harvest credentials, AD data, and BitLocker keys. CVE-2026-111357 and CVE-2026-118527 were added to the CISA KEV catalog Oct 30, 2026, with federal remediation deadlines of Nov 2 and Nov 13, 2026. Fixed in Ivanti Connect Secure 25.1.1.3 / 22.8R2.6. The Hacker News (source)
🟠 High impact

Langflow — critical unauthenticated RCE and auth-bypass flaws

CVE-2026-93680CVE-2026-110186LangflowCVSS 9.8

WhatLangflow 1.0.0 to 1.12.8 has unauthenticated auth-bypass and code-execution flaws; a Metasploit module exists but no in-the-wild exploitation is confirmed.

DoUpgrade Langflow to 1.13.0, which also disables the default auto-login as superuser.

Details and sources
IBM's advisory for the open-source AI workflow builder covers CVE-2026-93680 (CVSS 9.8, CWE-287 improper authentication, lets an unauthenticated attacker bypass security restrictions) and CVE-2026-110186 (CVSS 9.8, CWE-94 code injection, unauthenticated arbitrary code execution via MCP tool enumeration), plus CVE-2026-110193 (CVSS 8.8, OS command execution for a low-privileged account) and CVE-2026-110192 (CVSS 6.5, credential disclosure). All affect Langflow 1.0.0–1.12.8; fixed only in 1.13.0 (released Oct 22, 2026), which also turns off the previously on-by-default "auto-login as superuser" setting. A public Metasploit module exists for CVE-2026-93680 but currently only enumerates tools rather than executing code; not yet in CISA KEV, no confirmed in-the-wild exploitation. The Hacker News (source)
🟠 High impact

Canonical LXD / MicroCloud — maximum-severity certificate-impersonation flaw

CVE-2026-112816LXDMicroCloudIncusCVSS 10.0

WhatAn attacker with a trusted client's public certificate and network access to the LXD API can impersonate that client and control the cluster and VMs.

DoUpgrade to LXD 6.10.1/5.21.9/5.0.11/4.0.15, MicroCloud 3.4.1/2.1.5 or Incus 7.6.0, or restrict network access to the API.

Details and sources
CVE-2026-112816 (CVSS 10.0, CWE-295 improper certificate validation) lets an attacker with network access to the HTTPS API and a trusted client's public certificate impersonate that client — no private key required — gaining full cluster and VM access. Affects LXD 4.0.14/5.0.10/5.21.8/6.9 and earlier in each line, plus MicroCloud (the Incus fork shares the flaw). Fixed in LXD 6.10.1/5.21.9/5.0.11/4.0.15, MicroCloud 3.4.1/2.1.5, Incus 7.6.0; no workaround beyond restricting network access to the API. Not in CISA KEV, no known exploitation yet. The Hacker News (source)
🟠 High impact

Linux kernel TCP use-after-free (MPTCP remote root) — Red Hat now patched, others still exposed

Added after publishingCVE-2026-98432Linux kernel TCPCVSS 9.8

WhatA use-after-free in IPv4 TCP connection setup allows unauthenticated remote root, shown by Anthropic research; no in-the-wild exploitation is confirmed.

DoInstall Red Hat's Nov 1 kernels for RHEL 8.10, 9.8 and 10.2 and reboot; track Ubuntu's stable-channel fix.

Follow-ups: 17/11
Details and sources
CVE-2026-98432 (CVSS raised to 9.8/Critical, up from Red Hat's earlier 5.9) is a use-after-free in the IPv4 TCP connection-setup path (net/ipv4/af_inet.c/tcp_ipv4.c, present since kernel 4.4 — note the flaw is in plain TCP, not MPTCP itself, though Anthropic's proof-of-concept exploit targeted an MPTCP listener) that Anthropic's Claude used to build a working unauthenticated remote-root exploit, disclosed as research rather than an in-the-wild attack. Red Hat shipped fixed kernels Nov 1 for RHEL 8.10, 9.8, and 10.2 (reboot required); Debian, Oracle Linux, SUSE, and Amazon Linux fixed it in October. Ubuntu's fix is still only in the opt-in -proposed testing channel for 22.04/24.04/26.04, with 16.04/18.04/20.04 and several variants still pending — worth tracking for a stable-channel release. No confirmed in-the-wild exploitation; not in CISA KEV as of Oct 30. The Hacker News (source)

Breaches / Incidents 8

🟠 High impact

FBI attributes record $6.43B Binance crypto theft to North Korea's TraderTraitor

TraderTraitorBinance$6.43BNorth Korea

WhatUS agencies attributed the $6.43B Binance theft across 11 blockchains to North Korea's TraderTraitor; $2.85B has been frozen on-chain.

Follow-ups: 13/11
Details and sources
The FBI, U.S. Secret Service, and IRS-CI formally attributed the theft of $6.43B in virtual assets from Binance — the largest crypto theft on record — to the North Korea-aligned group TraderTraitor (aka Jade Sleet, UNC4899). Funds were taken across 11 blockchains; Tether, Circle, Ripple, Arbitrum, and Optimism have frozen $2.85B on-chain. No arrests reported yet. The Hacker News (source)
🟠 High impact

Mercado Bitcoin, Brazil's largest crypto exchange, hacked — ~$46M drained

Mercado BitcoinBrazilabout $46M

WhatBrazil's largest crypto exchange suspended deposits and withdrawals after an attack; researchers say over $46M left the platform, which the company disputes.

Details and sources
Mercado Bitcoin (~4.5M customers) suspended deposits/withdrawals after an attack discovered overnight Nov 1–2. Blockchain researchers say over $46M left the platform; the company disputes a "data breach" occurred and says customer funds weren't harmed, which several researchers dispute based on on-chain evidence. Tether and other platforms have frozen roughly $6.5M combined. The Record (source)
🟠 High impact

WorkOS breach via AI-assisted "ARTEX" framework exposes customer Slack/CRM data

WorkOSARTEXinternal sales app

WhatAn AI-assisted framework was used to access a WorkOS internal sales app, exposing CRM records and excerpts of customer Slack and sales call conversations.

Details and sources
Identity-infrastructure provider WorkOS confirmed unauthorized access (detected Oct 30) to an internal sales application, found to involve an AI-assisted offensive-security framework. Exposed data includes CRM records, customer account histories, and summaries/excerpts of sales calls, support discussions, and customer Slack conversations. WorkOS says no evidence production services were affected and is still determining which records were actually accessed. Cyber Security News (source)
🟠 High impact

Northflank breach — customers told to treat all secrets as stolen

NorthflankWizAI developer platforms

WhatNorthflank disclosed a breach and users report misuse of AI API keys, part of a wave hitting at least nine AI developer platforms.

DoNorthflank customers should treat every secret in environment variables as compromised and rotate it.

Details and sources
Cloud/AI dev platform Northflank disclosed a breach (incident started Oct 30, flagged by Wiz) and told customers to treat every secret in environment variables as compromised; users have reported misuse of AI API keys. Part of a broader October wave Wiz tracked across at least nine AI developer platforms, including Railway, OpenCode, Warp, and Lightning AI. Resultsense (source)
🔴 Exploited or active

New multi-country wave of supermarket refrigeration cyberattacks hits 8 European countries

Danfosssupermarket refrigeration8 countries

WhatOvernight into Nov 2, supermarket cooling systems, mostly Danfoss, in eight European countries were forced into defrost or had setpoints raised, spoiling food.

DoUpdate outdated controller software, move units off insecure networks and replace default passwords.

Details and sources
Overnight into Nov 2, refrigeration/cooling systems (mostly Danfoss) at supermarkets across Belgium, Sweden, Ireland, Estonia, the Netherlands, Finland, Portugal, and Italy were manipulated into defrost mode or had setpoints raised, spoiling food — a distinct, newly dated development beyond the original French incident. Named retailers affected include Delhaize, Carrefour, Intermarché, Spar, Musgrave/SuperValu, Selver, and Kesko. Danfoss controllers were hit in both this wave and the original French attack, and Danfoss says the common thread is outdated, unpatched software — the units hit in France were older models it no longer updates, and the affected Dutch systems were on insecure networks with default passwords — but Danfoss and The Hacker News are explicit that a shared vendor/method does not establish the same attackers carried out both waves. (This source does not mention Albert Heijn/Ahold Delhaize's separate October IT incident, which involves no refrigeration angle and remains unconnected to either refrigeration wave.) Belgium's federal prosecutor opened a criminal investigation; no group has claimed responsibility and Belgium's Centre for Cybersecurity says attackers "may never be determined." The Hacker News (source)
🟡 Worth knowing

St. Luke's Medical Center (Philippines) confirms data breach

St. Luke's Medical CenterPhilippinesthird-party platform

WhatSt. Luke's disclosed unauthorized access to a third-party platform supporting its mobile app, affecting an unspecified number of patients and employees.

Details and sources
St. Luke's disclosed unauthorized access to a third-party-managed platform supporting its mobile app, affecting an unspecified number of patients and employees. The hospital says the incident was contained immediately, operations were unaffected, and affected individuals will be notified directly. Daily Tribune (source)
🟡 Worth knowing

Online fashion retailer Revolve discloses card breach

Added after publishingRevolvesaved payment cardsSEC filing

WhatAn unauthorized actor obtained names, addresses and saved card numbers from Revolve; 127,200 Texans are confirmed and the 1.5 million figure is unconfirmed.

Details and sources
Revolve Group told customers and the SEC that an unauthorized actor obtained names, addresses, and saved payment card numbers/expiration dates, possibly "through automated means"; incident dated to Sept 11, discovered Sept 29. Only customers with a saved card were affected. The widely-reported figure of 1.5 million affected people is unconfirmed — it traces to a single Texas attorney-general filing, which separately confirms 127,200 affected Texans; Revolve's own SEC filing gives no total. No attacker has claimed responsibility. Malwarebytes (source)
🟡 Worth knowing

Group hijacks digital billboards in Egypt, Malaysia, and Brazil to display explicit content

Added after publishingLeakers$digital billboardsEgypt, Malaysia, Brazil

WhatLeakers$ claimed it pushed explicit content to billboards through exposed content-management software; Egypt blamed a compromise of the operator's website.

Details and sources
A group calling itself Leakers$ claimed responsibility, including a billboard overlooking a busy Cairo highway on Oct 29 that displayed the group's alias and a Telegram link. The group says it didn't hack individual screens but found the content-management software used to push content to displays exposed online, and that scanning turned up thousands of vulnerable systems in about two hours; Egyptian authorities instead attributed it to a compromise of the billboard operator's own website. Separate claims of access in Iraq and China are unconfirmed. Malwarebytes (source)

Malware / Ransomware 6

🔴 Exploited or active

BYOD ransomware gang hits a hospital, then "apologizes," and claims Michigan State University

BYODUNM HealthMichigan State University

WhatThe BYOD gang attacked UNM Health's hospital, then offered keys and payments to patients, and claims data on 1.3 million Michigan State students.

Details and sources
The BYOD gang (emerged Sept 2026) attacked University of New Mexico Health's hospital (ambulances diverted, surgeries suspended Oct 26), then said it hadn't realized some machines supported real-time patient care, offered decryption keys for hospital devices, and offered patients up to $2,500 each while keeping the stolen data — UNM says it made no payments and warns the offers may be scams. The same gang separately posted Michigan State University to its leak site, claiming theft of personal data on 1.3 million current/former students after an early-October credential-compromise incident; MSU has not confirmed the claim and says its central Active Directory shows no successful compromise. Daily Hodl (source) · The Record (source)
🔴 Exploited or active

C3RB3R ransomware hits Synology NAS devices via patched DSM flaw

C3RB3RSynology DSMSynology-SA-24:20

WhatSynology warns that attackers target unpatched DSM devices, and customers report C3RB3R ransomware appending .L0CK3D and dropping read-me3.txt.

DoUpdate Synology DSM to 7.2.2-72806-1 or the fixed version for your line.

Details and sources
Synology warns that threat actors are targeting devices still running outdated, unpatched DSM versions; customers report infection by C3RB3R ransomware, which appends a ".L0CK3D" extension and drops a "read-me3.txt" ransom note. Likely tied to the already-patched Synology-SA-24:20 advisory (CVE-2024-10441, CVE-2024-50629, CVE-2024-10445), fixed by Dec 2024 — the exact exploited flaw and scale of compromise aren't officially confirmed. Fixed DSM versions: 7.2.2-72806-1 and later per line. The Hacker News (source)
🟠 High impact

VOLTA infostealer uses Claude/Kimi/Gemini to rank stolen victim data by financial value

VOLTAinfostealerLLM victim ranking

WhatVOLTA steals browser credentials, cards, wallet seeds and session tokens, then uses Claude, Kimi or Gemini to rank victims before exfiltrating via Telegram.

Details and sources
Cisco Talos documented VOLTA (VOLTA-AUTODOXER), a Windows infostealer that harvests browser credentials, payment cards, crypto wallet seed phrases, and session tokens, then sends the haul via OpenRouter to Claude Opus 4.6 (falling back to Kimi K2, then Gemini 2.5 Flash) to generate a ranked victim "dossier" for the operator before exfiltrating via Telegram. The Hacker News (source)
🟡 Worth knowing

NetamLoader — new .NET loader hits a government Windows 10 build in Asia

NetamLoader.NET loaderAsian government

WhatKaspersky found a new AppDomainManager-injection loader delivered via a CAB file in an update directory on one government Windows 10 build.

Details and sources
Kaspersky disclosed NetamLoader, a previously undocumented AppDomainManager-injection loader delivered via a Windows CAB file placed in an update directory, used against a government-only Windows 10 build with a single known victim; a supply-chain delivery vector is suspected. The Hacker News (source)
🟠 High impact

Fake "IT support" callers on Microsoft Teams abuse AD CS to steal company data

Microsoft Teamsfake IT supportAD CS abuse

WhatAttackers posing as help desk on external Teams tenants got Quick Assist access, then abused misconfigured AD CS to mint privileged certificates and exfiltrate data.

DoFix insecure AD CS configurations such as ESC1 and ESC8 style misconfigurations.

Details and sources
Microsoft detailed two intrusions where attackers impersonated help-desk staff over external Teams tenants, talked victims into granting Quick Assist remote control, then abused insecurely configured Active Directory Certificate Services (ESC1/ESC8-style misconfigurations, not CVEs) to mint certificates for privileged accounts — including a domain controller in one case — and exfiltrate data via a renamed Rclone copy. The Hacker News (source)
🟠 High impact

Google disrupts 1.3-million-device Android botnet, accuses TV-box maker of selling criminal access

Videostrongx.cpe backdoor1.3M devices

WhatGoogle says TV-box maker Videostrong built a firmware backdoor and sold criminal access; Google seized the botnet's domains and servers.

Details and sources
Google says it's "highly confident" Shenzhen-based Android TV box maker Videostrong deliberately built a backdoor, dubbed x.cpe, into its firmware and sold criminal access to it — not an accident. The botnet topped 1.3M devices (mostly TV boxes, but also telehealth devices, karaoke machines, and digital signage) used for ad fraud, residential-proxy abuse, and credential/keystroke theft. Google seized the botnet's domains/servers with help from Japan's NPA, Lumen, HUMAN Security, Plume, Synthient, and TrendAI, and banned Videostrong from building certified Android devices. The Hacker News (source)

Threat Actor Activity 6

🟠 High impact

DarkBlinders (Iran-nexus) runs espionage campaign against an Israeli target and Iraqi Kurdish government cloud

DarkBlindersIran-nexusIsrael, Iraqi Kurdistan

WhatDarkBlinders used spear-phishing and a fake Dubai Airports recruiter lure to compromise an Israeli individual and a Kurdistan government cloud from August to October.

Details and sources
DarkBlinders (medium-confidence link to APT34) ran an Aug–Oct 2026 campaign using spear-phishing, credential-phishing lookalike domains, and a fake Dubai Airports recruiter lure to compromise a high-profile individual in Israel and a government cloud environment in Iraq's Kurdistan Region, using tools including ShelbyLoader V2 and a GitHub-based C2 backdoor. The Hacker News (source)
🟡 Worth knowing

SysGoRAT / CL-UNK-1211 — suspected AI-assisted espionage campaign against government Linux servers

SysGoRATCL-UNK-1211government Linux

WhatAn unattributed cluster targeted government Linux servers from October 2025 to April 2026 with a Go RAT and PHP web shells, some likely AI-generated.

Details and sources
Palo Alto Networks Unit 42 documented an unattributed cluster (CL-UNK-1211) running Oct 2025–Apr 2026 against Linux infrastructure tied to government and state-affiliated organizations, deploying a custom Go-based RAT, SysGoRAT, alongside PHP web shells to steal credentials and government mailboxes. Researchers assess some of the PHP/shell tooling was likely AI-generated — one script had syntax errors that broke its own execution. The Hacker News (source)
🟡 Worth knowing

Belarusian Cyber Partisans and Silent Crow claim destructive attack on Russia's Rosstat

Belarusian Cyber PartisansSilent CrowRosstat

WhatTwo groups claimed a destructive attack and data theft at Russia's statistics agency; Rosstat confirmed targeting but disputed the claimed scope.

Details and sources
Belarusian Cyber Partisans and pro-Ukraine group Silent Crow claimed a destructive cyberattack on Russia's federal statistics agency, Rosstat, and theft of sensitive government data. Rosstat confirmed it was targeted but disputed the claimed scope; the claims are not independently verified. The Record (source)
🟠 High impact

Hitcore Corp and DC8044 hijack Ukraine's and Russia's emergency-alert apps to spread fake nuclear warnings

Hitcore CorpDC8044emergency alert apps

WhatHackers pushed fake radiation, missile and nuclear alerts through Ukraine's Air Alert app and Russia's EMERCOM app on Nov 1.

Details and sources
Late Nov 1, hackers compromised Ukraine's widely used Air Alert app and pushed fake alerts about radiation, chemical threats, and a missile strike, with a group calling itself Hitcore Corp claiming responsibility; Ukraine's Air Force says the app was "likely compromised by the enemy" (unconfirmed) and disabled push notifications before restoring service. Hours later, a group calling itself DC8044 claimed it breached Russia's EMERCOM app via its API and sent false warnings that nuclear warheads had launched toward Russia; Russia's Emergency Situations Ministry confirmed the app was hacked but says its primary siren/broadcast warning system was unaffected. The Record (source)
🟡 Worth knowing

Apple sends mercenary-spyware threat notifications to users in 99 countries

Applethreat notifications99 countries

WhatApple sent a new round of mercenary-spyware threat notifications to users in 99 countries, without naming a vendor or nation.

Details and sources
A fresh round of Apple threat notifications warned targets across 99 countries of suspected mercenary-spyware targeting, with no vendor or nation attributed in this round; notified individuals reportedly include a Turkey-based reporter and a Lebanese documentary filmmaker. The Record (source)
🔴 Exploited or active

Russia-aligned Gorbag hackers target Ukraine exclusively, exploit patched WinRAR flaw

Added after publishingGorbagUkraineCVE-2025-8088WinRAR

WhatRussia-aligned Gorbag targets Ukrainian law enforcement, justice and defense with GIFTEDCROOK and BeautyStealer, in one case exploiting a patched WinRAR flaw.

Details and sources
ESET is tracking Gorbag (active since at least Feb 2025; linked to CERT-UA's UAC-0226 and Trend Micro's SHADOW-EARTH-066), which focuses exclusively on Ukrainian law-enforcement, justice, front-line local-government, and defense targets. It delivers the GIFTEDCROOK browser/credential stealer and BeautyStealer (which also grabs OpenVPN/RDP configs and KeePass databases) via government-themed spear-phishing, in one case exploiting the already-patched WinRAR flaw CVE-2025-8088 to trigger DLL sideloading. One malicious script contained a prompt-injection comment aimed at AI code reviewers claiming it wasn't malicious — testing showed AI assistants still flagged it correctly as a reverse shell. ESET suggests Gorbag may act as an initial-access broker for other actors, unconfirmed. The Hacker News (source)
‹ 02/11/202604/11/2026 ›