Threat Intelligence Brief - Tuesday, July 21, 2026

Threat Radar

• ServiceNow AI Platform (CVE-2026-6875, CVSS 9.5) is being actively exploited for unauthenticated RCE within days of disclosure — any unpatched enterprise instance is exposed now.

• Qilin ransomware is weaponizing a critical authentication bypass in Palo Alto PAN-OS GlobalProtect, turning perimeter VPN infrastructure into a direct ransomware entry point.

• WordPress wp2shell (CVE-2026-63030 / CVE-2026-60137) is under mass exploitation fueled by a public exploit — unauthenticated RCE and full site compromise are achievable at scale.

• JADEPUFFER has deployed purpose-built Go ransomware (ENCFORGE) against Langflow AI servers, specifically targeting model weights and vector indexes — a novel threat to AI/ML infrastructure.

• Clover Health disclosed a social engineering-driven breach of employee accounts exposing personal and health data, reinforcing that identity remains the dominant healthcare attack surface.

Immediate Action Required

• ServiceNow AI Platform — CVE-2026-6875 (CVSS 9.5): Confirm patch status across all ServiceNow AI Platform instances immediately. In-the-wild exploitation is confirmed. Core ITSM workflows and sensitive enterprise data are directly at risk. Patch or isolate before end of business today.

• Palo Alto PAN-OS GlobalProtect — Authentication Bypass: Verify that all GlobalProtect-exposed instances are patched and review perimeter logs for anomalous authentication activity. Qilin is actively using this flaw as a ransomware entry point. Engage network security and IT operations now.

• WordPress — CVE-2026-63030 / CVE-2026-60137 (wp2shell): Inventory all WordPress deployments. Mass scanning is underway and a public exploit is available. Unpatched sites face immediate takeover. Web operations and application owners should treat this as emergency patching.

• Langflow AI Servers — ENCFORGE Ransomware: If your organization runs Langflow in any environment, validate exposure, restrict external access, and verify backup integrity for model weights and vector indexes. JADEPUFFER has demonstrated repeat targeting of the same infrastructure.

High-Impact Developments

ServiceNow AI Platform CVE-2026-6875 Actively Exploited for Unauthenticated RCE

• What happened: A critical sandbox escape vulnerability (CVE-2026-6875, CVSS 9.5) in the ServiceNow AI Platform is being exploited in the wild for unauthenticated remote code execution. Exploitation was observed within days of public disclosure, confirmed by threat intelligence firm Defused Cyber.

• Why it matters: ServiceNow is deeply embedded in enterprise IT operations — ITSM workflows, HR processes, and sensitive operational data all flow through it. Unauthenticated RCE requires no credentials to achieve code execution on the platform. Exploitation within days of disclosure leaves a minimal patching window.

• Who should care: Security leadership, SOC, cloud operations, application owners, and vulnerability management teams responsible for ServiceNow environments.

• Recommended action: Apply the vendor patch immediately across all instances, including dev and staging. Review access logs for anomalous activity since the disclosure date. Escalate to leadership if same-day patching is not achievable.

• Confidence: High — confirmed in-the-wild exploitation reported by two independent sources.

• Search metadata: CVE-2026-6875, T1190, T1059, ServiceNow AI Platform, Sandbox Escape, Remote Code Execution

Intelligence Context

• Critical ServiceNow AI Platform Flaw Exploited for Unauthenticated Code Execution — The Hacker News
  • Context: Threat intelligence firm Defused Cyber confirmed in-the-wild exploitation of CVE-2026-6875 as a sandbox escape enabling unauthenticated code execution on the ServiceNow AI Platform.
• Exploitation of ServiceNow Vulnerability Seen Days After Disclosure — SecurityWeek
  • Context: SecurityWeek corroborated active exploitation of CVE-2026-6875 occurring within days of public disclosure, underscoring the compressed patching window for enterprise defenders.

Qilin Ransomware Exploiting Critical Palo Alto GlobalProtect Authentication Bypass

• What happened: The Qilin ransomware gang is actively exploiting a critical authentication bypass in Palo Alto Networks PAN-OS GlobalProtect to breach victim networks, according to Arctic Wolf. This converts a perimeter VPN appliance into a direct ransomware ingress point.

• Why it matters: GlobalProtect is a widely deployed enterprise VPN solution. An authentication bypass at the perimeter eliminates the need for credential theft or phishing — attackers move directly from internet to internal network. Qilin is an established ransomware-as-a-service operation with a documented history of double extortion.

• Who should care: Security leadership, SOC, network security teams, and IT operations managing Palo Alto infrastructure.

• Recommended action: Verify patch status on all PAN-OS GlobalProtect instances. Review authentication logs for anomalous or unauthenticated access attempts. If patching is delayed, assess whether temporary access restrictions or compensating controls reduce exposure. Confirm incident response readiness.

• Confidence: High — active exploitation confirmed by Arctic Wolf.

• Search metadata: T1190, Qilin, PAN-OS GlobalProtect, Palo Alto Networks, Ransomware, Authentication Bypass

Intelligence Context

• Critical Palo Alto VPN bug now exploited by Qilin ransomware gang — Bleeping Computer
  • Context: Arctic Wolf attributed active exploitation of the PAN-OS GlobalProtect authentication bypass to the Qilin ransomware gang, confirming this is a live ransomware campaign rather than opportunistic scanning.

WordPress wp2shell Mass Exploitation Underway

• What happened: Attackers are mass-exploiting two chained critical WordPress vulnerabilities — CVE-2026-63030 and CVE-2026-60137, collectively dubbed “wp2shell” — to achieve unauthenticated RCE and complete site compromise. A public exploit is available and is driving widespread automated scanning.

• Why it matters: A public exploit combined with mass scanning means the exploitation window is effectively immediate for any unpatched WordPress instance. Full site compromise enables defacement, data theft, malware hosting, and use of the site as an attack pivot against visitors or downstream systems.

• Who should care: SOC, web operations, application owners, and any team responsible for WordPress-based properties — including marketing, e-commerce, and customer-facing portals.

• Recommended action: Inventory all WordPress deployments across the organization, including shadow IT and agency-managed sites. Apply patches immediately. For sites that cannot be patched immediately, assess whether WAF rules or temporary takedown is appropriate given the public exploit availability.

• Confidence: High — active mass exploitation confirmed with public exploit in circulation.

• Search metadata: CVE-2026-63030, CVE-2026-60137, T1190, T1059, WordPress, wp2shell, Remote Code Execution

Intelligence Context

• WordPress wp2shell Exploitation Grows as Public Exploit Fuels Mass Scanning — The Hacker News
  • Context: Confirmed that the two chained WordPress flaws are under active mass exploitation, with a public exploit accelerating scanning and compromise attempts against vulnerable sites at scale.

ENCFORGE Ransomware Targets AI Infrastructure via Langflow RCE

• What happened: Threat actor JADEPUFFER — previously documented as an AI-agent-driven operator — has deployed ENCFORGE, a compiled Go-based ransomware, against Langflow AI servers. ENCFORGE is purpose-built to encrypt model weights, vector indexes, and related AI/ML assets. JADEPUFFER has been observed conducting repeat attacks against the same Langflow infrastructure, indicating persistent targeting.

• Why it matters: Ransomware designed specifically to destroy AI model assets can cause severe disruption to AI-dependent business processes, with recovery requiring retraining or restoration of large model files. This is not general-purpose ransomware repurposed against AI infrastructure — it is purpose-built for it.

• Who should care: Security leadership, SOC, cloud operations, and AI/ML platform owners running Langflow or similar AI orchestration platforms.

• Recommended action: Identify all Langflow deployments and confirm whether they are internet-exposed. Restrict access to trusted networks only. Verify that model weights and vector indexes are backed up to isolated, immutable storage. Review Sysdig’s JADEPUFFER research for indicators of compromise.

• Confidence: High — confirmed by Sysdig research with attributed threat actor and named malware.

• Search metadata: T1190, T1059, T1486, JADEPUFFER, ENCFORGE, Langflow, Ransomware, Remote Code Execution

Intelligence Context

• New ENCFORGE Ransomware Targets AI Model Files in Langflow RCE Attack — The Hacker News
  • Context: Sysdig researchers linked ENCFORGE deployment to JADEPUFFER, confirming a second attack on the same Langflow server and establishing this as a persistent, targeted campaign against AI infrastructure rather than an isolated incident.

Clover Health Data Breach via Social Engineering

• What happened: Clover Health disclosed that attackers used social engineering to compromise employee accounts, gaining access to personal and health information. No CVE or technical vulnerability was involved — the attack vector was human manipulation leading to account takeover.

• Why it matters: Healthcare data breaches carry significant HIPAA regulatory exposure, potential state privacy law liability, and patient trust consequences. Social engineering is the dominant initial access technique across the healthcare sector and is not fully addressable through technical controls alone.

• Who should care: Security leadership, privacy and legal/compliance teams, SOC, and identity and access management teams — particularly in healthcare and adjacent regulated industries.

• Recommended action: Review phishing and social engineering awareness training currency, MFA enforcement across all employee-facing systems, and privileged account access controls. Assess whether your organization’s employee account compromise detection capabilities would have identified this pattern.

• Confidence: High — disclosed by the organization directly, reported by SecurityWeek.

• Search metadata: T1566, T1078, Social Engineering, Data Breach, Healthcare

Intelligence Context

• Clover Health Investments Discloses Data Breach — SecurityWeek
  • Context: SecurityWeek reported Clover Health’s disclosure that social engineering was used to compromise employee accounts with access to personal and health information, confirming the breach vector and data types affected.

Monitor Only

• Zimbra vulnerabilities (command injection, XSS, SSRF) and a LegacyHive zero-day with privilege escalation (T1548) appear in scope for this brief cycle but no source articles were available at time of publication. Watch for additional reporting before actioning. Source: Global metadata context — no source article available.

Analyst Observation

This brief is dominated by a single pattern: critical vulnerabilities in widely deployed enterprise and internet-facing platforms exploited within days — sometimes hours — of disclosure. ServiceNow, GlobalProtect, and WordPress all share the same operational reality: the window between patch release and active exploitation has collapsed. Weekly or monthly patching cycles for internet-exposed infrastructure are no longer defensible.

The ENCFORGE story warrants separate attention. Ransomware operators are beginning to treat AI/ML infrastructure as a distinct, high-value target class. If your organization runs AI workloads on platforms like Langflow without the same security controls applied to production databases, that gap is now a liability.

The Clover Health breach is a reminder that none of the above technical controls matter if an attacker can social-engineer past them. Identity remains the soft underbelly across every sector.

Source Links

• Critical Palo Alto VPN bug now exploited by Qilin ransomware gang — https://www.bleepingcomputer.com/news/security/critical-globalprotect-vpn-bug-now-exploited-in-ransomware-attacks/

• WordPress wp2shell Exploitation Grows as Public Exploit Fuels Mass Scanning — https://thehackernews.com/2026/07/wordpress-wp2shell-exploitation-grows.html

• New ENCFORGE Ransomware Targets AI Model Files in Langflow RCE Attack — https://thehackernews.com/2026/07/new-encforge-ransomware-targets-ai.html

• Exploitation of ServiceNow Vulnerability Seen Days After Disclosure — https://www.securityweek.com/exploitation-of-servicenow-vulnerability-seen-days-after-disclosure/

• Critical ServiceNow AI Platform Flaw Exploited for Unauthenticated Code Execution — https://thehackernews.com/2026/07/critical-servicenow-ai-platform-flaw.html

• Clover Health Investments Discloses Data Breach — https://www.securityweek.com/clover-health-investments-discloses-data-breach/

Generated by DevSecOpsDadAttack cyber threat intelligence.