daily critical

CTI Daily Brief: 2026-05-02 — cPanel Zero-Day Mass-Exploited in Sorry Ransomware Campaign; Shinyhunters, M3rx, and Everest All Active

26 reports across 5 sources. CVE-2026-41940 cPanel auth-bypass exploited at scale to deploy Sorry ransomware on 44,000+ hosts. Critical libssh2 and binutils RCEs disclosed. Shinyhunters, M3rx, and Everest post fresh victims.

Reporting Period Classification Generated
2026-05-02 (24h) TLP:CLEAR 2026-05-03

1. Executive Summary

The pipeline processed 26 reports across 5 sources in the last 24 hours, with 3 critical and 12 high-severity items. The dominant story is confirmed in-the-wild mass-exploitation of cPanel CVE-2026-41940, an authentication-bypass flaw being weaponised to deploy the Go-based Linux “Sorry” ransomware; Shadowserver telemetry indicates at least 44,000 cPanel-running IP addresses have already been compromised. Two additional critical software flaws were disclosed by Microsoft: CVE-2026-7598 (libssh2 integer overflow in userauth_password, remote code execution path) and CVE-2026-6846 (binutils arbitrary code execution via malformed XCOFF object files). Ransomware leak-site activity remained heavy, with Shinyhunters posting Instructure Holdings (Canva LMS) and Cushman & Wakefield, the M3rx group posting four new manufacturing/IT/HVAC victims, and Everest naming Fiserv. Microsoft Defender briefly false-positive-flagged legitimate DigiCert root certificates as Trojan:Win32/Cerdigent.A!dha before issuing a corrective signature update, an incident plausibly linked to the recent DigiCert code-signing breach.

2. Severity Distribution

Severity Count Key Drivers
🔴 CRITICAL 3 cPanel CVE-2026-41940 in-the-wild RCE; libssh2 CVE-2026-7598; binutils CVE-2026-6846
🟠 HIGH 12 Shinyhunters (Instructure, Cushman & Wakefield); M3rx (4 victims); Everest (Fiserv); Telegram FEMITBOT scam platform; libpng/Hex/Perl Storable/binutils CVEs
🟡 MEDIUM 4 Microsoft Defender DigiCert false-positive; nano CVEs (-6842, -6843); vidtv CVE-2026-43058
🟢 LOW 2 Wireshark 4.6.5 (43 vulns / 38 CVEs patched); CVE-2026-30656 disclosure
🔵 INFO 5 Telegram OSINT chatter (Darkfeed pulse, BreachForums council); Photonic listing on mnt6

3. Priority Intelligence Items

3.1 cPanel CVE-2026-41940 Mass-Exploited to Deploy “Sorry” Ransomware

Source: BleepingComputer

A critical authentication-bypass flaw in cPanel and WHM tracked as CVE-2026-41940 is being mass-exploited as a zero-day. Shadowserver reports at least 44,000 cPanel-running IP addresses have been compromised since the emergency patch window opened, with exploitation activity dating back to late February. Successful exploitation grants attackers control-panel access; operators are then dropping a Go-based Linux encryptor for the “Sorry” ransomware family, which appends .sorry to encrypted files, uses ChaCha20 file encryption with an embedded RSA-2048 public key, and instructs victims to negotiate via a Tox ID in a README.md ransom note. Researcher Rivitna assesses decryption is impossible without the corresponding RSA-2048 private key. Affected products: cPanel, WHM (Linux web hosting control panels). Affected sectors: hosting providers, SMB websites, e-commerce. ATT&CK: T1190 (Exploit Public-Facing Application), T1486 (Data Encrypted for Impact), T1203 (Exploitation for Client Execution).

Indicators of Compromise

Encrypted file extension: .sorry
Ransom note filename: README.md
Tox ID: 3D7889AEC00F2325E1A3FBC0ACA4E521670497F11E47FDE13EADE8FED3144B5EB56D6B198724
Encryption: ChaCha20 + RSA-2048
Family: Sorry (Go-based Linux encryptor — unrelated to 2018 HiddenTear .sorry campaign)

SOC Action: Patch all cPanel/WHM instances to the emergency hotfix release immediately. For internet-facing cPanel hosts, hunt for unauthorised process execution under the cPanel service account, the presence of README.md ransom notes in web roots, files with .sorry extension, and outbound Tor connectivity. Block the published Tox ID at perimeter and EDR. Treat any cPanel host that was internet-exposed since late February 2026 as potentially compromised — review web-server access logs for anomalous authenticated WHM sessions and create a backup-restore plan in case of encryption.

3.2 CVE-2026-7598 — libssh2 Integer Overflow in userauth_password (Remote Code Execution)

Source: Microsoft MSRC

Microsoft disclosed a critical integer-overflow vulnerability in libssh2’s userauth.c::userauth_password routine. Improper handling of large integer values during password authentication allows buffer overflow conditions, creating a remote code-execution path against any application that links the affected libssh2 versions and exposes SSH client functionality to untrusted input. Affected products: any product bundling libssh2 prior to the patched release (commonly used by file-transfer agents, Git clients, monitoring tools, embedded SSH clients). ATT&CK: T1068 (Exploitation for Privilege Escalation), T1078 (Valid Accounts).

SOC Action: Inventory all software dependent on libssh2 (use SBOM data or ldd/grep sweeps across Linux estate; on Windows, search for bundled libssh2.dll). Prioritise patching of any agent that initiates outbound SSH to externally controlled hosts. Until patches are applied, restrict outbound SSH from server estate to a known-good destination allowlist and alert on libssh2-using processes spawning unexpected child processes.

3.3 CVE-2026-6846 — Binutils Arbitrary Code Execution via Malformed XCOFF Files (paired with CVE-2026-6845 DoS and CVE-2025-11083 ELF Heap Overflow)

Source: Microsoft MSRC — CVE-2026-6846, CVE-2026-6845, CVE-2025-11083

A trio of GNU Binutils vulnerabilities were disclosed in the same window, correlated by the pipeline (confidence 0.80, infrastructure correlation): CVE-2026-6846 enables arbitrary code execution when binutils processes malformed XCOFF object files; CVE-2026-6845 causes a denial-of-service via crafted ELF; CVE-2025-11083 is a heap-based overflow in elf_swap_shdr within elfcode.h. The combined risk is meaningful for any CI/CD pipeline, malware reverse-engineering workstation, or build host that processes untrusted object files. ATT&CK: T1059 (Command and Scripting Interpreter), T1203.

SOC Action: Update binutils packages across all developer workstations, build servers, and forensic/malware-analysis hosts. Ensure analyst sandboxes processing third-party binaries are isolated and snapshot-restorable. Audit CI runners for jobs that invoke objdump, readelf, nm, or ld against artifacts originating from public registries or pull requests.

3.4 Telegram “FEMITBOT” Mini-App Platform Drives Crypto Scams and Android Malware

Source: BleepingComputer

CTM360 researchers exposed FEMITBOT, a large-scale fraud platform abusing Telegram’s Mini App feature. Threat actors operate a shared backend (signature API response: “Welcome to join the FEMITBOT platform”) behind multiple phishing domains and Telegram bots impersonating Apple, Coca-Cola, Disney, eBay, IBM, Moon Pay, NVIDIA, and YouKu. When a victim clicks “Start” on the bot, the embedded Mini App renders a phishing page in Telegram’s WebView, displays fake balances and countdown timers, then either solicits deposits/referrals or pushes Android APKs hosted on the same TLS-validated domain as the API. APK names mimic legitimate apps such as BBC, NVIDIA, CineTV, Coreweave, and Claro. Tracking pixels (Meta, TikTok) are embedded for conversion optimisation. ATT&CK: T1566 (Phishing), T1189 (Drive-by Compromise), T1124 (System Information Discovery).

Indicators of Compromise

Platform identifier (API response): "Welcome to join the FEMITBOT platform"
Delivery vector: Telegram Mini Apps + bot Start button
Payload type: Android APKs sideloaded outside Google Play
Brands abused: Apple, Coca-Cola, Disney, eBay, IBM, Moon Pay, NVIDIA, YouKu, BBC, CineTV, Coreweave, Claro

SOC Action: Add a detection in proxy/EDR for the FEMITBOT API string in HTTP responses. For corporate-issued Android devices, enforce MDM policy blocking sideloading and unknown sources. Brief finance and treasury teams on Telegram-hosted “investment” lures. Threat-hunt for Telegram WebView-originated traffic to newly registered domains carrying brand-impersonation TLS certs.

3.5 Ransomware Leak-Site Surge — Shinyhunters, M3rx, and Everest

Source: Shinyhunters / RansomLook, M3rx / RansomLook, Everest / RansomLook

Three ransomware groups posted fresh activity in the reporting window, drawing a high-confidence “multi-sector targeting” trend from the AI correlation layer (batch 102, confidence 0.90 actor correlation on M3rx). Shinyhunters named Instructure Holdings, Inc. (Canva LMS / instructure.com — education sector) and Cushman & Wakefield Inc. (commercial real estate); the group continues to operate from shinyhunte.rs and rotating onion infrastructure, with phishing as the documented initial-access vector. M3rx added four victims in 24 hours: Engineered Machine Tool Inc. (manufacturing — Wichita KS, 180 GB / 698k files claimed stolen), Freitag IT GmbH (IT solutions, Germany), Manatee Air Heating & Cooling Inc. (HVAC, Florida), with prior-week posts including Boxtopia (UK packaging, 166 GB), Optimization Software Technologies LLP (India, 222 GB), DM Schweiz (Switzerland, 120 GB), and Anvil Arts (UK performing arts). Everest named Fiserv (financial-services technology — payments processing). Everest’s broader recent leak-site cadence shows continued double-extortion and pure data-extortion activity, including TSYS, Epiq Global, Symcor, Liberty Mutual, and Frost Bank in late April. ATT&CK: T1566 (Phishing), T1078 (Valid Accounts), T1486 (Data Encrypted for Impact), T1496 (Resource Hijacking).

SOC Action: For organisations in the named verticals (education LMS, commercial real estate, manufacturing, IT services, HVAC, financial-services tech), assume opportunistic targeting and run an immediate phishing-readiness check: validate MFA enforcement on all SSO entry points, confirm M365/Google Workspace mailbox-rule monitoring is on, and verify that domain-controller audit logging plus offline backups are current. For Fiserv-adjacent customers, monitor for fraud-pattern shifts and review any third-party integrations sharing data with Fiserv.

3.6 Microsoft Defender False-Positive Flags Legitimate DigiCert Root Certificates

Source: BleepingComputer

Microsoft Defender signature update (initial release 30 April) flagged two legitimate DigiCert root certificate thumbprints as Trojan:Win32/Cerdigent.A!dha, in some cases removing them from the Windows AuthRoot trust store under HKLM\SOFTWARE\Microsoft\SystemCertificates\AuthRoot\Certificates\. Affected admins reported broken TLS validation and unnecessary OS reinstalls. Microsoft fixed the detection in Security Intelligence build 1.449.430.0 (current 1.449.431.0), which also restores the removed entries. The incident appears plausibly linked to the recent DigiCert code-signing breach in which attackers used initialisation codes against approved-but-undelivered EV code-signing certificate orders, leading to revocation of 60 certificates (27 connected to “Zhong Stealer” malware). ATT&CK: T1566 (Phishing — initial DigiCert support staff compromise vector).

Indicators of Compromise

Detection (false-positive): Trojan:Win32/Cerdigent.A!dha
Affected DigiCert root thumbprints (legitimate, do NOT block):
  0563B8630D62D75ABBC8AB1E4BDFB5A899B24D43
  DDFB16CD4931C973A2037D3FC83A4D7D775D05E4
Fixed in Defender Security Intelligence: 1.449.430.0 (or later)
Registry path affected: HKLM\SOFTWARE\Microsoft\SystemCertificates\AuthRoot\Certificates\

SOC Action: Confirm all managed Windows endpoints have Defender signature 1.449.430.0 or later — push via WSUS/Intune and validate. For systems that experienced the false positive, verify the two DigiCert root thumbprints are restored to AuthRoot. Separately, audit code-signing certificate-issuance webhooks and revocation feeds for DigiCert-issued certs since early April; treat any internally observed signed binary with revoked DigiCert EV certs as suspect (Zhong Stealer association).

Risk Trend Supporting Evidence
🔴 CRITICAL Exploitation of critical vulnerabilities in widely used software CVE-2026-7598 libssh2 RCE; CVE-2026-41940 cPanel “Sorry” ransomware mass-exploitation
🟠 HIGH Multi-sector ransomware targeting using overlapping TTPs Cushman & Wakefield (Shinyhunters); emtco.com, it-freitag.de, manateeair.com (M3rx); Fiserv (Everest)
🟠 HIGH Increased exploitation of software vulnerabilities across different sectors CVE-2026-6845 binutils DoS; CVE-2026-32148 Hex lockfile checksum bypass
🟡 MEDIUM Phishing as a prevalent TTP across diverse cyber threats Defender / DigiCert incident; Telegram FEMITBOT Mini Apps; Shinyhunters Instructure breach; CVE-2026-32148 Hex dependency-substitution risk

Top correlation entries from batch 102/103: actor correlation on M3rx with confidence 0.90 (manufacturing, IT solutions, HVAC); infrastructure correlation across binutils CVE pair at 0.80; T1059 TTP correlation between nano (-6842) and binutils (-6846) at 0.80; T1486 correlation between Shinyhunters Cushman & Wakefield post and Sorry ransomware campaign at 0.70.

Threat Actors

  • Qilin (82 reports, 30-day) — ransomware-as-a-service operator, last seen 2026-05-02
  • The Gentlemen (63 reports, 30-day) — sustained leak-site activity through April
  • Coinbase Cartel (31 reports) — financially motivated, last seen 2026-04-23
  • DragonForce (27 reports) — ransomware affiliate ecosystem, last seen 2026-04-22
  • shadowbyt3$ (25 reports) — extortion-focused actor on leak markets
  • ShinyHunters / Shinyhunters (37 reports combined across casing variants) — last seen 2026-05-03 (Instructure Holdings)
  • Inc Ransom (16 reports) — last seen 2026-05-02
  • M3rx (3 mentions in this period) — newer multi-sector RaaS posting four victims today
  • Everest (1 mention this period; 455 leak-site posts all-time) — pivoted toward pure data extortion
  • FEMITBOT (this period) — Telegram Mini App fraud operator

Malware Families

  • RansomLook / RansomLock (97 reports combined) — ransomware leak-site monitoring tag dominant in feed
  • RaaS (23 reports) — generic RaaS-tagged activity
  • Tox1 / Tox (34 reports combined) — Tox-based negotiation infrastructure across multiple groups
  • Qilin (11 reports as malware family)
  • Gentlemen ransomware (9 reports)
  • Sorry (1 report this period — new entrant tied to cPanel CVE-2026-41940)
  • Trojan:Win32/Cerdigent.A!dha (1 report — Defender false-positive label, not a real family)

6. Source Distribution

Source Reports URL Notes
Microsoft 11 link All vulnerability-alert items: 3 CVEs in binutils, libssh2 critical, Hex/libpng/Perl/nano/vidtv mid-tier disclosures
RansomLook 7 link Leak-site coverage of Shinyhunters (×2), M3rx (×3), Everest (×1), mnt6 (×1)
Unknown 4 All Telegram-origin OSINT (Darkfeed pulse + BreachForums council discussion). Telegram URLs withheld per editorial policy
BleepingComputer 3 link Headline coverage: cPanel Sorry ransomware, Telegram FEMITBOT, Defender DigiCert false-positive
SANS 1 link Wireshark 4.6.5 release (43 vulns / 38 CVEs / 35 bugs — high count attributed to AI-assisted vuln reporting)

7. Consolidated Recommendations

  • 🔴 IMMEDIATE: Patch cPanel/WHM to the emergency build addressing CVE-2026-41940 and threat-hunt for Sorry ransomware indicators (.sorry files, README.md ransom notes, the published Tox ID, outbound Tor connectivity from cPanel hosts). Treat any cPanel host internet-exposed since late February 2026 as potentially compromised.
  • 🔴 IMMEDIATE: Inventory and update libssh2-linked software (CVE-2026-7598). Prioritise outbound-SSH agents and any service that processes attacker-controlled SSH credentials.
  • 🟠 SHORT-TERM: Roll the binutils trio (CVE-2026-6846 / -6845 / CVE-2025-11083) onto developer workstations, build servers, and malware-analysis sandboxes. Audit CI pipelines for object-file processing of untrusted artifacts.
  • 🟠 SHORT-TERM: Confirm Microsoft Defender signature 1.449.430.0+ deployed across managed estate and verify the two DigiCert root thumbprints are restored to AuthRoot on previously affected hosts. Separately review any internally observed code-signed binaries against the DigiCert revocation list (Zhong Stealer association).
  • 🟡 AWARENESS: Brief finance, M&A, and brand-protection teams on the FEMITBOT Telegram Mini-App scam pattern. Education, commercial real estate, manufacturing, IT services, HVAC, and financial-services-technology verticals should treat this week’s Shinyhunters / M3rx / Everest leak-site posts as evidence of opportunistic multi-sector pressure and verify MFA, mailbox-rule monitoring, and offline backup status.
  • 🟢 STRATEGIC: The pipeline’s repeated correlation on T1566 (Phishing) across vulnerability disclosures, ransomware leak posts, and OSINT-tracked operations underlines that phishing remains the single most operationally significant initial-access TTP this week. Consider an organisation-wide tabletop on phishing-driven supply-chain and code-signing compromise (the DigiCert support-staff incident is a worked example).

This brief was generated entirely by AI from automated threat intelligence collection and correlation pipelines, made up of 26 reports processed across 2 correlation batches. A human analyst reviewed and approved this report before publication, but AI-generated analysis may contain errors in attribution, severity assessment, or indicator extraction. Always verify IOCs, CVE details, and threat actor attribution against primary sources before taking operational action.