<?xml version="1.0" encoding="UTF-8"?>
<threatroundup feed="roundup" title="ThreatRoundup: The Roundup" publisher="ThreatRoundup.com" generated="2026-10-29T21:10:19+00:00" page="1">
  <description>Every ThreatRoundup story: what happened, what it means for you, and what defenders should do.</description>
  <link>https://threatroundup.com/news/</link>
  <item>
    <title>Revoked the M365 Session? Attackers Can Still Read Mail for 90 Minutes</title>
    <link>https://threatroundup.com/news/m365-revoke-sessions-graph-ews-90-minutes/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://labs.eye.security/how-attackers-keep-reading-mail-after-m365-session-revocation/">Eye Security</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>4</huntLeads>
    <cves>0</cves>
    <summary>Eye Security found that revoking sessions or resetting a password leaves Graph, EWS and ActiveSync mail access open for up to 90 minutes; only disabling the account cuts it off immediately.</summary>
  </item>
  <item>
    <title>A 16-Year-Old UEFI Bug Lets One Packet Beat Secure Boot</title>
    <link>https://threatroundup.com/news/edk2-pxe-uefi-flaw-cve-2026-97631/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/critical-16-year-old-uefi-flaw-can-run.html">The Hacker News</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>1</cves>
    <summary>CVE-2026-97631, a 9.5-rated overflow in EDK II&apos;s PXE driver dating to 2010, can let someone on the local network run code before Secure Boot on network-booting machines. Firmware updates are still pending.</summary>
  </item>
  <item>
    <title>Sony’s So-net Breach Spills 2.6 Million ISP Customer Records</title>
    <link>https://threatroundup.com/news/sony-so-net-isp-breach-2-5-million/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/security/sony-isp-subsidiary-data-breach-exposes-25-million-customer-records/">BleepingComputer</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Sony Network Communications says an Oct. 19 intrusion exposed 2.59 million So-net and Docomo Hikari customer records, plus 54,401 email addresses with passwords. No card data was taken.</summary>
  </item>
  <item>
    <title>Fake Card-Protection Calls Plant Bank Trojans on 18,000 Indian Phones</title>
    <link>https://threatroundup.com/news/cyberdakait-android-banking-trojans-india/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.group-ib.com/blog/cyberdakait-seven-android-trojans-indian-banking/">Group-IB</source>
    <threatLevel>High</threatLevel>
    <huntLeads>5</huntLeads>
    <cves>0</cves>
    <summary>Group-IB tracked seven Android trojan families behind fake bank &apos;protection plan&apos; calls in India: nearly 18,000 phones infected and 7.6 million texts, including OTPs, intercepted.</summary>
  </item>
  <item>
    <title>SkyW4lker Phishing Asks for a Code, Not Your Microsoft 365 Password</title>
    <link>https://threatroundup.com/news/skyw4lker-device-code-phishing-m365/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://any.run/cybersecurity-blog/skyw4lker-phishing-kit-analysis/">ANY.RUN</source>
    <threatLevel>High</threatLevel>
    <huntLeads>5</huntLeads>
    <cves>0</cves>
    <summary>ANY.RUN tracked the SkyW4lker kit across 173 domains: password-protected lures send Microsoft 365 users to the real device login page, where one typed code hands the attacker a token.</summary>
  </item>
  <item>
    <title>Inside Qilin: The Password-Locked Encryptor That Takes Down ESXi Too</title>
    <link>https://threatroundup.com/news/qilin-ransomware-acronis-esxi-deep-dive/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.acronis.com/en/tru/posts/inside-qilin-ransomware-esxi-attacks-affiliate-networks-and-double-extortion/">Acronis TRU</source>
    <threatLevel>High</threatLevel>
    <huntLeads>6</huntLeads>
    <cves>0</cves>
    <summary>Acronis TRU dissects Qilin&apos;s Rust encryptor, its password gate and the PowerShell routine that resets ESXi root passwords and encrypts every host, as the gang passes 900 claimed victims in 2026.</summary>
  </item>
  <item>
    <title>Fake ‘GPT-6’ and Claude Apps Hide a Miner Behind Real Signed Installers</title>
    <link>https://threatroundup.com/news/yoba-fake-chatgpt-claude-installers-sideloading/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://research.checkpoint.com/2026/tricks-from-the-past-deployed-via-signed-chatgpt-and-claude-binaries/">Check Point Research</source>
    <threatLevel>High</threatLevel>
    <huntLeads>6</huntLeads>
    <cves>0</cves>
    <summary>Check Point&apos;s Yoba campaign pairs genuine signed OpenAI and Anthropic installers with a sideloaded DLL on fake GitHub AI-app pages, installing a hidden XMRig miner via a Telegram-seeded DGA.</summary>
  </item>
  <item>
    <title>China-Linked Hackers Used a VMware ESX Bug to Work Without Logs</title>
    <link>https://threatroundup.com/news/vmware-esx-logging-zero-day-warp-panda-kev/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/china-linked-hackers-exploited-vmware.html">The Hacker News</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>1</cves>
    <summary>CrowdStrike says China-nexus Warp Panda used CVE-2026-41709, a VMware ESX insufficient-logging flaw, as a zero-day against US law firms. CISA added it to KEV with a Nov. 11 deadline.</summary>
  </item>
  <item>
    <title>Volt Typhoon’s ‘Silent Train’ Rode Four Zero-Days Into Network Control</title>
    <link>https://threatroundup.com/news/volt-typhoon-silent-train-zero-days-lumen/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.lumen.com/blog/en-us/silent-train-volt-typhoon-operational-framework">Lumen Black Lotus Labs</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>6</huntLeads>
    <cves>1</cves>
    <summary>Lumen ties Volt Typhoon&apos;s Silent Train covert network to zero-days in N-able, Cisco SD-WAN, VeloCloud and likely Quest KACE, while Cisco Talos details KeyDrift malware stealing SD-WAN keys.</summary>
  </item>
  <item>
    <title>Cisco Firewall VPN Bug Is Being Used to Knock ASA and FTD Offline</title>
    <link>https://threatroundup.com/news/cisco-asa-ftd-vpn-dos-cve-2026-76418/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/cisco-patches-actively-exploited-asa.html">The Hacker News</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>1</cves>
    <summary>Cisco says attackers are exploiting CVE-2026-76418, an unauthenticated bug in the ASA and FTD SSL VPN that forces the firewall to reboot when secondary authentication is configured.</summary>
  </item>
  <item>
    <title>Magento Loader on 953 Stores Shows Scanners a Blank Page</title>
    <link>https://threatroundup.com/news/magento-cms-block-loader-953-stores/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://sansec.io/research/magento-cms-block-loader-deepveindrilling">Sansec</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>2</cves>
    <summary>Sansec found a JavaScript loader in Magento CMS blocks on 953 stores since July 2025; the newest version fingerprints visitors so scanners see nothing, and PolyShell is the main way in.</summary>
  </item>
  <item>
    <title>Aokigahara Stealer Grabs Your Claude, Cursor and Copilot Sessions</title>
    <link>https://threatroundup.com/news/aokigahara-infostealer-ai-coding-tools/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://research.checkpoint.com/2026/infostealer-riding-the-ai-train/">Check Point Research</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>6</huntLeads>
    <cves>0</cves>
    <summary>Check Point says the new Aokigahara infostealer has a dedicated module for AI assistant and coding-tool sessions, plus cloud configs, SSH keys, browsers and crypto wallets.</summary>
  </item>
  <item>
    <title>ENGIE Customer Data Stolen Through a Partner’s Sales Tool</title>
    <link>https://threatroundup.com/news/engie-france-customer-data-partner-marketing-tool/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/security/energy-giant-engie-discloses-data-breach-affecting-customers/">BleepingComputer</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>ENGIE says attackers stole French customers&apos; contact details and contract and meter numbers from a partner marketing platform; no banking data, and its own systems weren&apos;t hit.</summary>
  </item>
  <item>
    <title>DreamHost Locks Out Email and FTP Users, Won’t Say Why</title>
    <link>https://threatroundup.com/news/dreamhost-forced-password-resets-incident/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/security/dreamhost-forces-email-ftp-sftp-password-resets-in-ongoing-incident/">BleepingComputer</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>DreamHost is forcing email, FTP and SFTP password resets in an ongoing incident and telling owners to remove unknown forwarding rules, but hasn&apos;t said what happened.</summary>
  </item>
  <item>
    <title>Inbox Flood, Then ‘IT’ on Teams: 400 Cases End in Remote Control</title>
    <link>https://threatroundup.com/news/teams-fake-it-support-quick-assist-ransomware/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://arcticwolf.com/resources/blog/threat-actors-abuse-microsoft-teams-external-chat-for-initial-access/">Arctic Wolf</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Arctic Wolf has seen more than 400 incidents where attackers mail-bomb a user, pose as IT on Microsoft Teams and get remote access via Quick Assist, a path to Black Basta, Cactus and Chaos ransomware.</summary>
  </item>
  <item>
    <title>Hidden Behind NAT, Still Hijacked: Dahua Cameras Hit in Ukraine</title>
    <link>https://threatroundup.com/news/dahua-cameras-hijacked-ukraine-p2p-cert-ua/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/attackers-hijack-dahua-ip-cameras-in.html">The Hacker News</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>2</cves>
    <summary>CERT-UA warns attackers are hijacking Dahua cameras and recorders in Ukraine through Easy4IP P2P access, guessable serials, weak passwords and old auth-bypass bugs.</summary>
  </item>
  <item>
    <title>Australia’s Spies Unplug an Amadey Botnet From 180,000 PCs</title>
    <link>https://threatroundup.com/news/asd-amadey-botnet-180000-disconnected/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/australia-gets-into-amadey-malware.html">The Hacker News</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Australia&apos;s ASD broke into an Amadey malware control panel and disconnected it from about 180,000 infected computers. Data already stolen is still out there.</summary>
  </item>
  <item>
    <title>Cyberattack Sends New Mexico’s Only Top Trauma Center Into Diversion</title>
    <link>https://threatroundup.com/news/unm-health-cyberattack-ambulance-diversion/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.securityweek.com/cybersecurity-issue-forces-a-major-new-mexico-hospital-to-divert-patients-and-close-clinics/">SecurityWeek (AP)</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>A cybersecurity incident forced UNM Health to divert ambulances from New Mexico&apos;s only Level 1 trauma center and cancel appointments at more than 30 clinics. The FBI is assisting.</summary>
  </item>
  <item>
    <title>Hacker Trades in Victims’ Accounts, Then Tries to Extort GMO</title>
    <link>https://threatroundup.com/news/gmo-onamae-desktop-cloud-trading-accounts-extortion/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://ir.group.gmo/en/pdf/irlibrary/gmo_disclose_info20261024_e.pdf">GMO Internet Group</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>GMO Internet says an attacker who breached Onamae.com Desktop Cloud used stolen logins to trade in customers&apos; brokerage accounts and is now demanding money. GMO won&apos;t pay.</summary>
  </item>
  <item>
    <title>Apple Brings Its Exploited CoreGraphics Fix to Older iPhones</title>
    <link>https://threatroundup.com/news/apple-ios-18-7-11-coregraphics-zero-day-backport/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/apple/apple-backports-coregraphics-zero-day-patches-to-older-iphones/">BleepingComputer</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>1</cves>
    <summary>iOS and iPadOS 18.7.11 backport Apple&apos;s fix for CVE-2026-86950, a CoreGraphics zero-day Apple says may have been exploited against targeted individuals, to the iPhone XS, XR and older iPads.</summary>
  </item>
  <item>
    <title>FBI: Any Campaign Donation That ‘Pays You Back’ Is a Scam</title>
    <link>https://threatroundup.com/news/fbi-midterm-campaign-donation-scams/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://moneywise.com/news/top-stories/fbi-warning-midterm-election-scams-campaign-fraud">MoneyWise</source>
    <threatLevel>Info</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>The FBI warns of midterm scams promising guaranteed returns on campaign donations, fake PACs that keep the money, merchandise that never ships and fake voter-registration pages.</summary>
  </item>
  <item>
    <title>Tailworm: Tailscale SSH Bugs Can Turn One Login Into Root</title>
    <link>https://threatroundup.com/news/tailworm-tailscale-ssh-root-flaws/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://tailworm.com/">Cyberdione Labs (Tailworm)</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Cyberdione Labs&apos; Tailworm report says flaws in Tailscale SSH&apos;s root-run forwarding let a permitted SSH user reach root-only sockets. Tailscale calls the issues serious and targets a patched client by Oct. 27.</summary>
  </item>
  <item>
    <title>Gheist Turns Stolen GitHub Tokens Into Secret Dumps via Actions Logs</title>
    <link>https://threatroundup.com/news/gheist-github-tokens-actions-logs-secrets/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.wiz.io/blog/gheist-attempted-scaled-abuse-of-stolen-github-tokens-stumbles">Wiz</source>
    <threatLevel>High</threatLevel>
    <huntLeads>6</huntLeads>
    <cves>0</cves>
    <summary>Wiz says the Gheist campaign used stolen GitHub tokens to plant 13,500+ branches across 850+ accounts, each running a workflow that dumps repository secrets into Actions logs and deletes the run.</summary>
  </item>
  <item>
    <title>Stolen Login Tokens Opened 97 Microsoft 365 Accounts, No Password Needed</title>
    <link>https://threatroundup.com/news/m365-token-replay-97-accounts-huntress/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.huntress.com/blog/token-replay-spray">Huntress</source>
    <threatLevel>High</threatLevel>
    <huntLeads>6</huntLeads>
    <cves>0</cves>
    <summary>Huntress tracked a campaign that replayed stolen Microsoft 365 sign-in tokens to take over 97 accounts at 79 organizations, then planted rogue Entra devices, swapped MFA methods and hid payroll mail.</summary>
  </item>
  <item>
    <title>Fortinet Says Patch Now, Explain Later: Critical Fixes, No CVEs Until Nov. 10</title>
    <link>https://threatroundup.com/news/fortinet-ai-found-flaws-patch-before-cves/</link>
    <published>2026-10-29T12:45:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.cyberdaily.au/security/14348-act-now-acsc-warns-of-critical-vulnerabilities-in-fortios-fortimanager-fortianalyzer-and-fortimail">Cyber Daily</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Fortinet patched critical flaws in FortiOS, FortiManager, FortiAnalyzer, FortiMail and FortiAuthenticator found by AI-assisted review, but won&apos;t publish CVEs until Nov. 10. Australia&apos;s ACSC says patch urgently.</summary>
  </item>
  <item>
    <title>Nefilim and Hive Renter Gets 70 Months in US Prison</title>
    <link>https://threatroundup.com/news/nefilim-hive-affiliate-stryzhak-70-months/</link>
    <published>2026-10-28T22:28:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/security/nefilim-and-hive-ransomware-affiliate-gets-70-months-in-prison/">BleepingComputer</source>
    <threatLevel>Info</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Artem Stryzhak, a Ukrainian affiliate who rented Nefilim and Hive ransomware to hit companies, law firms, a dental office and a healthcare provider, was sentenced to 70 months in US prison.</summary>
  </item>
  <item>
    <title>Visit the Wrong Site While Logged In to WordPress, Get a Backdoor</title>
    <link>https://threatroundup.com/news/trust-graft-wordpress-admin-plugin-backdoor/</link>
    <published>2026-10-28T22:28:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.securonix.com/blog/trust-graft-wordpress-admin-php-backdoor/">Securonix</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>5</huntLeads>
    <cves>0</cves>
    <summary>Securonix found JavaScript on hijacked sites that uses a logged-in WordPress admin&apos;s session to install a backdoored plugin that persists in mu-plugins and hides from the plugin list.</summary>
  </item>
  <item>
    <title>Leaked Chats Show Silent Ransom Group Squeezing Law Firms by Phone</title>
    <link>https://threatroundup.com/news/silent-ransom-group-leak-law-firms/</link>
    <published>2026-10-28T22:28:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://ransom-isac.org/blog/silent-ransom-group-leak-analysis/">Ransom-ISAC</source>
    <threatLevel>High</threatLevel>
    <huntLeads>4</huntLeads>
    <cves>0</cves>
    <summary>Ransom-ISAC&apos;s analysis of leaked Silent Ransom Group files shows callback-phishing extortion of mostly US law firms via AnyDesk, with at least $43.95 million confirmed received.</summary>
  </item>
  <item>
    <title>Brazilian Skimmers Rode an Analytics Script Into Dozens of Checkouts</title>
    <link>https://threatroundup.com/news/convertr-plausible-magecart-checkout-skimmer/</link>
    <published>2026-10-28T22:28:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://sansec.io/research/convertr-plausible-supply-chain-attack">Sansec</source>
    <threatLevel>High</threatLevel>
    <huntLeads>4</huntLeads>
    <cves>1</cves>
    <summary>Sansec says a Brazilian Magecart group planted a card skimmer in the Plausible analytics script every Convertr store loads, exposing checkouts at dozens of fashion brands Oct. 24 to 27.</summary>
  </item>
  <item>
    <title>Any Android App Could Ask Microsoft Authenticator for Your Work Tokens</title>
    <link>https://threatroundup.com/news/microsoft-authenticator-android-authmask-token-theft/</link>
    <published>2026-10-28T22:28:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://enclave.ai/blog/authmask-we-made-microsoft-authenticator-hand-over-tokens">Enclave</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>1</cves>
    <summary>Enclave&apos;s AuthMask bug let any Android app get work-account Microsoft Graph tokens from Authenticator by posing as Teams; Microsoft fixed it in 6.2606.4246 and says it wasn&apos;t exploited.</summary>
  </item>
  <item>
    <title>CISA Puts PolyShell and Node2Shell on Its Must-Patch List</title>
    <link>https://threatroundup.com/news/cisa-kev-polyshell-node2shell-exploit-traffic/</link>
    <published>2026-10-28T22:28:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/adobe-commerce-polyshell-attacks-breach.html">The Hacker News</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>2</cves>
    <summary>CISA added Adobe Commerce&apos;s PolyShell and the Brotli Node2Shell bug to its exploited-vulnerabilities list, due Nov. 10, as ELLIO logged the first Node2Shell exploit traffic.</summary>
  </item>
  <item>
    <title>PHP2Shell Scanners Left Their Toolboxes Open, and Hunt.io Looked</title>
    <link>https://threatroundup.com/news/php2shell-hunt-io-open-directory-iocs/</link>
    <published>2026-10-27T15:08:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://hunt.io/blog/php2shell-cve-2026-109618-109619-open-directories">Hunt.io</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>2</cves>
    <summary>Hunt.io combed about ten attacker open directories full of PHP2Shell tooling and a Mirai-style botnet, and published attempt signatures and indicators defenders can search for today.</summary>
  </item>
  <item>
    <title>A Korean-Speaking Crew Is Building a PHP2Shell Hit List</title>
    <link>https://threatroundup.com/news/cyfirma-dprk-php2shell-staging/</link>
    <published>2026-10-27T15:07:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.cyfirma.com/research/cyfirma-threat-advisory-suspected-dprk-campaign-staging-php2shell-targeting-against-global-php-infrastructure/">CYFIRMA</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>2</cves>
    <summary>CYFIRMA says Korean-speaking operators are quietly scanning 216 unpatched PHP hosts at governments, banks and logistics firms, with a moderate-confidence Lazarus link.</summary>
  </item>
  <item>
    <title>Oracle’s Emergency Java Fix Cuts eMudhra’s Weak Certificates Loose</title>
    <link>https://threatroundup.com/news/oracle-java-emergency-update-emudhra-cve-2026-91403/</link>
    <published>2026-10-27T15:06:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://cybersecuritynews.com/emergency-java-update/">Cyber Security News</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>2</cves>
    <summary>Oracle&apos;s out-of-band CVE-2026-91403 fix (CVSS 9.1) stops Java trusting eight eMudhra intermediate certificates with weak keys; a second update, CVE-2026-114820, removes the related roots.</summary>
  </item>
  <item>
    <title>k0t Walks In Through Fortinet and GlobalProtect, Leaves With 70,000 Logins</title>
    <link>https://threatroundup.com/news/k0t-network-edge-intrusions-group-ib/</link>
    <published>2026-10-27T15:05:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.group-ib.com/blog/stray-cat-network-edge/">Group-IB</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>5</cves>
    <summary>Group-IB and NyxLab say actor k0t exploited exposed FortiWeb, FortiGate and GlobalProtect devices through 2026, reached domain admin, and stole more than 70,000 credentials, likely as an initial access broker.</summary>
  </item>
  <item>
    <title>Node2Shell: Crafted Brotli Data Runs Code on a Stock Node.js Build</title>
    <link>https://threatroundup.com/news/node2shell-nodejs-brotli-rce-demo/</link>
    <published>2026-10-27T15:04:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/researchers-demonstrate-node2shell-rce.html">The Hacker News</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>1</cves>
    <summary>Imperva ran attacker code on an official Node.js 24.22.0 image by feeding it crafted Brotli data, through fetch() responses or Express 5 body parsers. Patch Node.js and rebuild images.</summary>
  </item>
  <item>
    <title>BlackX Bootkit Beats Secure Boot on Fully Patched Windows</title>
    <link>https://threatroundup.com/news/blackx-uefi-bootkit-baton-drop/</link>
    <published>2026-10-27T15:03:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://securelist.com/uefi-bootkit-blackx/121539/">Kaspersky Securelist</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>1</cves>
    <summary>Kaspersky found BlackX, a UEFI bootkit used against Central Asian diplomats that bypasses Secure Boot via Baton Drop because Microsoft&apos;s revocation step was never applied. Medium-confidence link to APT41.</summary>
  </item>
  <item>
    <title>Quest Diagnostics Breach Hits Its Lab Ordering Platform</title>
    <link>https://threatroundup.com/news/quest-diagnostics-patient-records-breach/</link>
    <published>2026-10-27T15:02:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.securityweek.com/quest-diagnostics-reports-cyber-breach-affecting-certain-patient-records/">SecurityWeek (Associated Press)</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Quest Diagnostics says an intruder reached patient records through its Quanum Lab Services Manager platform. It hasn&apos;t said how many people are affected.</summary>
  </item>
  <item>
    <title>Storm-0940’s AI Agents Move On to Okta, Entra and Cognito</title>
    <link>https://threatroundup.com/news/wiz-artex-storm-0940-expands-targeting/</link>
    <published>2026-10-27T15:01:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.wiz.io/blog/artex-in-the-wild-storm-0940-expands-targeting-of-it-and-software-vendors">Wiz Research</source>
    <threatLevel>High</threatLevel>
    <huntLeads>6</huntLeads>
    <cves>0</cves>
    <summary>Wiz says ARTEX AI agents tied to Storm-0940 spent October probing IT and software vendors&apos; Okta, Entra ID and Cognito sign-ins. Most attempts failed.</summary>
  </item>
  <item>
    <title>Researchers Defused a Popcorn Time Updater on Millions of PCs</title>
    <link>https://threatroundup.com/news/popcorn-time-updater-rce-disabled/</link>
    <published>2026-10-27T15:00:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/security/abandoned-popcorn-time-app-exposed-millions-of-device-to-rce-attacks/">BleepingComputer</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>1</cves>
    <summary>Eye Security found Time4Popcorn&apos;s abandoned updater would run any file as SYSTEM or root, registered its domain and used it to switch the updater off. Android users remain exposed.</summary>
  </item>
  <item>
    <title>Loblaw Says Its March Breach Hit 2 Million Shoppers</title>
    <link>https://threatroundup.com/news/loblaw-breach-2-million-credit-monitoring/</link>
    <published>2026-10-27T14:59:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.cbc.ca/news/business/loblaw-data-breach-apology-9.7393626">CBC News</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Loblaw&apos;s CEO has apologized to about two million customers whose names, addresses, birthdates and purchase details were taken in a February breach, and is offering two years of TransUnion monitoring.</summary>
  </item>
  <item>
    <title>Asus Router Bugs Let Neighbors Skip the Admin Password</title>
    <link>https://threatroundup.com/news/asus-router-firmware-cve-2026-93489/</link>
    <published>2026-10-27T14:58:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.notebookcheck.net/Asus-rushes-out-router-firmware-update-to-fix-critical-vulnerabilities.1433020.0.html">Notebookcheck</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>3</cves>
    <summary>Asus fixed three High-rated router flaws, including CVE-2026-93489 (CVSS 8.7), that let someone on your network act as admin without a password. No attacks reported yet.</summary>
  </item>
  <item>
    <title>One Unrevoked Key Let Attackers Hijack LiteLLM’s Release Pipeline</title>
    <link>https://threatroundup.com/news/litellm-leaked-master-key-release-hijack/</link>
    <published>2026-10-27T14:57:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/leaked-litellm-master-key-let-attackers.html">The Hacker News</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>LiteLLM says a master key leaked on GitHub in May and never revoked let an attacker steal maintainer tokens and ship a backdoored 1.107.0 to PyPI, Docker Hub and GHCR.</summary>
  </item>
  <item>
    <title>A Fake Amazon Ad Ends in finger.exe, Lumma and a Hijacked PC</title>
    <link>https://threatroundup.com/news/amazon-clickfix-finger-malware/</link>
    <published>2026-10-27T14:56:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.huntress.com/blog/amazon-clickfix-finger">Huntress</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Huntress says a sponsored Google result led to a fake Amazon login and a ClickFix CAPTCHA that used finger.exe to install the Lumma infostealer and NetSupport remote access.</summary>
  </item>
  <item>
    <title>KeyStealth Hid in a Microsoft 365 Inbox for Seven Weeks, MFA and All</title>
    <link>https://threatroundup.com/news/keystealth-device-code-phishing-kit/</link>
    <published>2026-10-27T14:55:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://adamnet.works/blog/keystealth-device-code-phishing-to-rogue-entra-devices-and-bec/">ADAMnetworks</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>ADAMnetworks details how a device code phishing kit let an operator register rogue Entra devices, sit in a company&apos;s Microsoft 365 account for seven weeks and attempt payment fraud.</summary>
  </item>
  <item>
    <title>Iran’s Prince of Persia Spies Move Their C2 Onto US Soil</title>
    <link>https://threatroundup.com/news/prince-of-persia-iran-apt-us-c2/</link>
    <published>2026-10-27T14:54:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.semperis.com/blog/prince-of-persia-threat-group-hiding-in-plain-sight/">Semperis</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Semperis says Iran-linked Prince of Persia (Infy) rebuilt its infrastructure after a February exposure and is now running a Tornado C2 server in the US for the first time in a decade.</summary>
  </item>
  <item>
    <title>Fake IT Support on Teams Delivers EtherRAT, C2 Via Ethereum</title>
    <link>https://threatroundup.com/news/etherrat-teams-fake-it-support-ethereum-c2/</link>
    <published>2026-10-27T14:53:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://bluecyber.hashnode.dev/etherrat-malware-bootcamp2026">bluecyber.hashnode.dev</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>6</huntLeads>
    <cves>0</cves>
    <summary>A Teams caller posing as IT talks victims into installing remote-access tools, then drops EtherRAT, a Node.js backdoor that looks up its command server in an Ethereum smart contract.</summary>
  </item>
  <item>
    <title>Old-School VBScript Lures Plant RATs Across Brazil and Portugal</title>
    <link>https://threatroundup.com/news/vbscript-malware-brazil-portugal-trusted-tools/</link>
    <published>2026-10-27T14:52:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.watchguard.com/wgrd-security-hub/secplicity-blog/how-vbscript-malware-hides-behind-trusted-tools-evade-detection">WatchGuard</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>5</huntLeads>
    <cves>0</cves>
    <summary>WatchGuard details three Brazil-linked campaigns, Overlord RAT, Lampion and a newdev.dll RAT, that start with VBScript and lean on built-in Windows tools and signed programs to stay hidden.</summary>
  </item>
  <item>
    <title>A Bayview Staffer Pasted 3,182 Patients’ Records Into an AI Chatbot</title>
    <link>https://threatroundup.com/news/johns-hopkins-bayview-patient-data-ai-tool/</link>
    <published>2026-10-27T14:51:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://hoodline.com/2026/10/johns-hopkins-bayview-warns-3-182-patients-after-staffer-fed-data-to-ai-tool/">Hoodline</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Johns Hopkins Bayview is notifying 3,182 patients after an employee uploaded a spreadsheet of their records, including medications and record numbers, to an unapproved public AI tool.</summary>
  </item>
  <item>
    <title>PHP2Shell Was Being Exploited Four Days Before Anyone Knew</title>
    <link>https://threatroundup.com/news/php2shell-exploited-before-disclosure-madpot/</link>
    <published>2026-10-22T11:31:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://aws.amazon.com/blogs/security/amazon-threat-intelligence-identifies-php2shell-exploitation-prior-to-public-disclosure-in-madpot/">Amazon Threat Intelligence (AWS Security Blog)</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>7</huntLeads>
    <cves>2</cves>
    <summary>Amazon says a PHP2Shell exploit hit its MadPot honeypots on Oct. 17, four days before PHP disclosed CVE-2026-109618 and CVE-2026-109619. It now calls exploitation broad.</summary>
  </item>
  <item>
    <title>Microsoft: China’s Storm-0940 Let AI Agents Run the ConnectWise Hack</title>
    <link>https://threatroundup.com/news/storm-0940-artex-connectwise-supply-chain/</link>
    <published>2026-10-22T11:30:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.microsoft.com/en-us/security/blog/2026/10/21/storm-0940-uses-artex-agentic-ai-connectwise-supply-chain-compromise/">Microsoft Threat Intelligence</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>7</huntLeads>
    <cves>0</cves>
    <summary>Microsoft ties the ConnectWise Cloud Backup supply-chain attack to China-aligned Storm-0940, which used ARTEX AI agents to try sign-ins to more than 2,250 Microsoft 365 tenants at up to 2,900 a minute.</summary>
  </item>
  <item>
    <title>GitLab’s CVSS 10 Regex Bug Reaches Back 12 Years to Version 7.0</title>
    <link>https://threatroundup.com/news/gitlab-cvss-10-unauthenticated-rce/</link>
    <published>2026-10-22T11:29:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/critical-gitlab-flaw-cvss-100-could.html">The Hacker News</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>4</cves>
    <summary>GitLab fixed CVE-2026-110264, a CVSS 10.0 flaw in its bundled Ruby regex engine that could let unauthenticated attackers run code on self-managed servers as old as version 7.0. No exploitation seen yet, and no workaround.</summary>
  </item>
  <item>
    <title>SafePay Ransomware Hit 600 Victims Using Stolen VPN Logins and Old Bugs</title>
    <link>https://threatroundup.com/news/cisa-safepay-ransomware-advisory-aa26-293a/</link>
    <published>2026-10-22T11:28:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.cisa.gov/news-events/cybersecurity-advisories/aa26-293a">CISA</source>
    <threatLevel>High</threatLevel>
    <huntLeads>9</huntLeads>
    <cves>5</cves>
    <summary>A joint FBI, CISA and USSS advisory says SafePay ransomware has hit at least 600 victims since August 2024, getting in through stolen or guessed VPN logins and old Cisco ASA, SonicWall and Veeam flaws, then wiping backups before it encrypts.</summary>
  </item>
  <item>
    <title>Node.js Patches a Brotli Bug That Squeezes Past Its Own Buffer</title>
    <link>https://threatroundup.com/news/nodejs-brotli-decoder-rce/</link>
    <published>2026-10-22T11:27:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/nodejs-brotli-decoder-bug-could-allow.html">The Hacker News</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>10</cves>
    <summary>Node.js fixed an out-of-bounds write in its bundled Brotli decoder (CVE-2026-111497) that it says could allow remote code execution on 22.x, 24.x and 26.x. End-of-life versions get no patch.</summary>
  </item>
  <item>
    <title>Lazarus Turns a Korean Banking Plug-In Into a Spy’s Front Door</title>
    <link>https://threatroundup.com/news/lazarus-operation-flashhole/</link>
    <published>2026-10-22T11:26:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://securelist.com/lazarus-operation-flashhole/121541/">Kaspersky Securelist</source>
    <threatLevel>High</threatLevel>
    <huntLeads>9</huntLeads>
    <cves>0</cves>
    <summary>Kaspersky says Lazarus abused South Korea&apos;s mandatory INISAFE CrossWeb EX security plug-in to break into at least three firms, using fake résumé sites, GitHub-based C2 and a post-quantum key exchange.</summary>
  </item>
  <item>
    <title>PaperCut Attacker Left an AI Hacking Workbench Wide Open</title>
    <link>https://threatroundup.com/news/papercut-attacker-ai-toolkit-second-rce/</link>
    <published>2026-10-22T11:25:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://hunt.io/blog/papercut-attacker-ai-toolkit-exploit-code-second-rce">Hunt.io</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>2</cves>
    <summary>Hunt.io found the PaperCut mass-exploitation crew&apos;s AI agent prompts, exploit variants, credential stealers and victim access kits on an open server, plus a second RCE route through the XML-RPC API.</summary>
  </item>
  <item>
    <title>Fake Cloudflare Check Hides Its C2 List on the Polygon Blockchain</title>
    <link>https://threatroundup.com/news/slaiclind-clickfix-etherhiding/</link>
    <published>2026-10-22T11:24:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.stormshield.com/news/cti-payload-slaiclind-campaign/">Stormshield</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>7</huntLeads>
    <cves>0</cves>
    <summary>Stormshield says hacked WordPress sites read C2 addresses from a Polygon smart contract, then show Windows visitors a fake Cloudflare check that leads to a stealer loaded through the Deno runtime.</summary>
  </item>
  <item>
    <title>Belarus-Aligned Hackers Hid a Backdoor in VMware Tools for 2 Years</title>
    <link>https://threatroundup.com/news/vasilek-backdoor-vmware-tools-russian-healthcare/</link>
    <published>2026-10-22T11:23:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/hackers-hid-vasilek-backdoor-in-vmware.html">The Hacker News</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Solar 4RAYS says Partisan Zmiy spent about two years in a Russian healthcare network, hiding the Telegram-controlled Vasilek backdoor and a fake service in an unmonitored VMware Tools directory.</summary>
  </item>
  <item>
    <title>Sandworm Gets a Handoff, Wipes a Ukrainian Grain Firm</title>
    <link>https://threatroundup.com/news/russian-hackers-wipe-ukrainian-grain-firm/</link>
    <published>2026-10-22T11:22:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://therecord.media/ukraine-hackers-russia-grain-sandworm">The Record</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>6</huntLeads>
    <cves>0</cves>
    <summary>ESET says UAC-0099 phished into a Ukrainian grain logistics company and handed access to Sandworm, which wiped data, while a new Go backdoor hid its C&amp;C inside the Yggdrasil mesh network.</summary>
  </item>
  <item>
    <title>Serengeti III: Interpol Rounds Up 1,462 Cybercrime Suspects in Africa</title>
    <link>https://threatroundup.com/news/africa-cybercrime-networks-disrupted/</link>
    <published>2026-10-22T11:21:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/security/authorities-disrupt-nearly-11-000-cybercrime-networks-in-africa/">BleepingComputer</source>
    <threatLevel>Medium</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>An Interpol-led operation across 21 countries arrested 1,462 suspects, dismantled 10,909 malicious networks and recovered more than $105 million from BEC, scam and ransomware crews, BleepingComputer reports.</summary>
  </item>
  <item>
    <title>Chrome Now Checks Your Clipboard Before ClickFix Scams Can Bite</title>
    <link>https://threatroundup.com/news/chrome-gemini-scam-detection-clickfix/</link>
    <published>2026-10-22T11:20:00+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/google/google-chrome-adds-gemini-powered-scam-detection-warns-users-of-clickfix-attacks/">BleepingComputer</source>
    <threatLevel>Info</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Chrome&apos;s Enhanced Protection now checks copied commands for ClickFix scams, uses server-side Gemini to spot scam pages, and adds a softer suspicious-site warning.</summary>
  </item>
  <item>
    <title>Qilin and INC Ransomware Slip Past MFA Through Unpatched PAN-OS</title>
    <link>https://threatroundup.com/news/pan-os-cve-2026-0257-qilin-inc-mfa-bypass/</link>
    <published>2026-10-21T12:47:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/qilin-and-inc-ransomware-groups-exploit.html">The Hacker News</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>2</cves>
    <summary>Unit 42 says the Qilin and INC ransomware groups are using CVE-2026-0257, a GlobalProtect authentication bypass patched in May, to open VPN sessions without MFA and then move on to vCenter, ESXi and Veeam.</summary>
  </item>
  <item>
    <title>Plex Servers Are Quietly Mining Crypto for Someone Else</title>
    <link>https://threatroundup.com/news/plex-media-server-exploited-cryptominer/</link>
    <published>2026-10-21T12:46:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.boredhackerblog.info/2026/10/plex-vulnerabilities-being-exploited-in.html">boredhackerblog</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>4</cves>
    <summary>Attackers are chaining four Plex Media Server bugs, now on CISA&apos;s exploited list, to drop a cryptominer and a traffic-relay proxy on servers that never got September&apos;s fix.</summary>
  </item>
  <item>
    <title>NetScaler Crash Handler Turned Into a Root Shell, Rapid7 Finds</title>
    <link>https://threatroundup.com/news/netscaler-cve-2026-88771-88779-chained-intrusions/</link>
    <published>2026-10-21T12:45:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.rapid7.com/blog/post/ra-netscaler-intrusions-chaining-cve-2026-88771-and-cve-2026-88779">Rapid7</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>2</cves>
    <summary>Rapid7 details intrusions where attackers chained CVE-2026-88771 and CVE-2026-88779 to get root on NetScaler ADC and Gateway, then planted HA-aware persistence and stole the appliance&apos;s credentials and keys.</summary>
  </item>
  <item>
    <title>Leaked ‘php2shell’ Bugs Force PHP Into an Emergency Patch Day</title>
    <link>https://threatroundup.com/news/php-cve-2026-109618-109619-critical-rce/</link>
    <published>2026-10-21T12:44:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/security/php-warns-admins-of-two-critical-flaws-emergency-patches-coming-tomorrow/">BleepingComputer</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>2</cves>
    <summary>PHP shipped 8.5.12, 8.4.27, 8.3.36 and 8.2.35 on Oct. 21 to fix two critical bugs nicknamed php2shell, after details leaked early. Updated Oct. 25: Amazon says both bugs are now broadly exploited.</summary>
  </item>
  <item>
    <title>GlassWorm Slips a Secret Stealer Into Five OVHcloud CLI Releases</title>
    <link>https://threatroundup.com/news/ovhcloud-cli-glassworm-backdoor/</link>
    <published>2026-10-21T12:43:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://opensourcemalware.com/blog/glassworm-hacks-ovh-clouds-cli-tool">OpenSourceMalware</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>An attacker used a hijacked developer account to swap the binaries on five OVHcloud CLI releases for a build that steals cloud keys, SSH keys and AI-tool data. OVHcloud ties it to GlassWorm and says about 250 downloads happened in the 94-minute window.</summary>
  </item>
  <item>
    <title>Rank Math PRO Shipped a Backdoor for 11 Hours</title>
    <link>https://threatroundup.com/news/rank-math-seo-plugin-backdoor/</link>
    <published>2026-10-21T12:42:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/security/wordpress-seo-plugin-rank-math-compromised-to-backdoor-websites/">BleepingComputer</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Rank Math PRO 3.0.123, served from the vendor&apos;s own site for about 11 hours on Oct. 15-16, carried a backdoor that creates hidden admins and rebuilds itself. Updating alone doesn&apos;t clean it.</summary>
  </item>
  <item>
    <title>A Stolen ConnectWise Credential Walked Into 173 Microsoft 365 Tenants</title>
    <link>https://threatroundup.com/news/connectwise-cloud-backup-supply-chain-attack/</link>
    <published>2026-10-21T12:41:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.truesec.com/hub/blog/supply-chain-attack-on-connectwise-cloud-backup">Truesec</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>An attacker used ConnectWise Cloud Backup&apos;s own app credential to sign in to 173 partner and customer Microsoft 365 tenants, per Huntress. ConnectWise has taken the service offline and disabled seven app identities.</summary>
  </item>
  <item>
    <title>APT28 Spent the Summer Inside a NATO Naval Shipbuilder</title>
    <link>https://threatroundup.com/news/apt28-nato-naval-shipbuilder-reflective-loader/</link>
    <published>2026-10-21T12:40:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://thehackernews.com/2026/10/russian-apt28-breaches-nato-naval.html">The Hacker News</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Symantec says Russia&apos;s APT28 breached a naval shipbuilder, a law enforcement agency and an aircraft-engine maker in two NATO countries, using the HOOKEDGE backdoor and a previously unseen reflective loader.</summary>
  </item>
  <item>
    <title>China-Linked Spies Turned BeyondTrust Help Desks Into Skeleton Keys</title>
    <link>https://threatroundup.com/news/armory-castle-beyondtrust-exploitation/</link>
    <published>2026-10-21T12:39:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://cloud.google.com/blog/topics/threat-intelligence/armory-castle-exploits-beyondtrust/">Google Threat Intelligence Group / Mandiant</source>
    <threatLevel>Critical</threatLevel>
    <huntLeads>8</huntLeads>
    <cves>1</cves>
    <summary>Google and Mandiant say suspected Chinese espionage group ARMORY CASTLE has exploited a BeyondTrust Remote Support zero-day since late August. Once inside, it took root, dumped stored credentials and used the support tool to reach internal machines.</summary>
  </item>
  <item>
    <title>Food Lion and Stop &amp; Shop Owner Still Digging Out From Cyberattack</title>
    <link>https://threatroundup.com/news/ahold-delhaize-cyberattack-us-europe/</link>
    <published>2026-10-21T12:38:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://www.bleepingcomputer.com/news/security/ahold-delhaize-cyberattack-affects-services-across-us-and-europe/">BleepingComputer</source>
    <threatLevel>High</threatLevel>
    <huntLeads>0</huntLeads>
    <cves>0</cves>
    <summary>Ahold Delhaize, parent of Food Lion, Stop &amp; Shop, Hannaford and Giant, is restoring websites and apps after a cyberattack disclosed Oct. 12. Stores are open and no data theft has been found so far.</summary>
  </item>
  <item>
    <title>AI-Written Malware Spent Six Months Inside Government Linux Servers</title>
    <link>https://threatroundup.com/news/unit42-ai-espionage-government-linux/</link>
    <published>2026-10-21T12:37:44+00:00</published>
    <publisher>ThreatRoundup.com</publisher>
    <source url="https://unit42.paloaltonetworks.com/ai-campaign-targets-linux-infrastructure/">Palo Alto Networks Unit 42</source>
    <threatLevel>High</threatLevel>
    <huntLeads>12</huntLeads>
    <cves>2</cves>
    <summary>Unit 42 says a cluster it calls CL-UNK-1211 spent six months in government Linux servers using AI-generated scripts, a new Go RAT and fake red-team markers, going after databases, mailboxes and linked social media accounts.</summary>
  </item>
</threatroundup>
