Audience: Vulnerability Managers, Security Operations, CISOs, DevSecOps Teams
Reading Time: Approximately 10 minutes

This Week's KEV Additions

Analysis

CVE-2026-73570 — Synacor / Zimbra Collaboration Suite (ZCS)

• What it is: An unauthenticated remote command injection flaw allowing attackers to execute OS commands via malicious SMTP requests when SNMP notifications are enabled.

• Affected versions: Zimbra Collaboration (ZCS) versions prior to 10.1.20 with the zimbra-snmp package.

• Exploitation status: Active in the wild (CISA KEV catalog).

• Patch available: Yes — vendor patch released in version 10.1.20.

• CISA due date: September 4, 2026.

• Operational risk: Critical perimeter vulnerability. Any internet-exposed Zimbra mail instance accepting raw SMTP traffic allows external threat actors to gain initial foot-level execution without user interaction.

CVE-2026-72530 — TrueConf / TrueConf Server

• What it is: A code injection flaw enabling network-adjacent or remote attackers to break out of isolated environments and achieve full server code execution.

• Affected versions: TrueConf Server versions 5.3.X to 5.3.9, 5.4.X to 5.4.9, and 5.5.X to 5.5.5.

• Exploitation status: Active in the wild — targeted campaigns weaponizing PhantomCore implant [Kaspersky / Securelist].

• Patch available: Yes — upgrade to TrueConf Server 5.5.6+.

• CISA due date: September 3, 2026.

• Operational risk: High impact on on-premises video infrastructure. Attackers reaching TCP port 4307 can break host containment and pivot directly into internal corporate networks.

CVE-2026-72529 — TrueConf / TrueConf Server

• What it is: Missing authentication on critical functions that allows unauthenticated remote script execution over administrative ports.

• Affected versions: TrueConf Server versions prior to 5.5.6.

• Exploitation status: Active in the wild (CISA KEV catalog).

• Patch available: Yes — vendor security release 5.5.6+.

• CISA due date: September 3, 2026.

• Operational risk: Frequently chained directly with CVE-2026-72530. Unauthenticated access on port 4307 removes all authorization barriers before initiating host environment breakouts.

CVE-2026-64849 — MLflow / MLflow

• What it is: An unauthenticated Server-Side Request Forgery (SSRF) flaw in webhook endpoints that reads back internal network and cloud metadata responses.

• Affected versions: MLflow versions prior to 3.15.0.

• Exploitation status: Active in the wild (CISA KEV catalog).

• Patch available: Yes — resolved in MLflow 3.15.0.

• CISA due date: September 2, 2026.

• Operational risk: Extreme risk to cloud-hosted AI workloads. Attackers can leverage the SSRF vector to query Instance Metadata Services (IMDS) and extract transient cloud IAM credentials.

CVE-2026-33824 — Microsoft / IKE Service Extensions

• What it is: A double-free memory corruption vulnerability in Windows IKE Extensions leading to unauthenticated remote code execution.

• Affected versions: Windows 10/11 and Windows Server 2016 through 2025.

• Exploitation status: Active in the wild — high public EPSS probability (~78%).

• Patch available: Yes — Microsoft Patch Tuesday security update.

• CISA due date: September 1, 2026.

• Operational risk: Critical threat to remote access infrastructure. Exposed IPsec/VPN gateways running unpatched Windows IKE services can be fully compromised via unauthenticated network packets.

CVE-2026-59310 — Broadcom / VMware vCenter

• What it is: A path traversal weakness allowing network-authenticated or adjacent attackers to execute arbitrary code on vCenter instances.

• Affected versions: Consult Broadcom Security Advisory for specific vCenter Server and Cloud Foundation build mappings.

• Exploitation status: Active in the wild (CISA KEV catalog).

• Patch available: Yes — official Broadcom security updates.

• CISA due date: September 1, 2026.

• Operational risk: Severe post-initial-access vector. Attackers who gain a baseline foothold can weaponize vCenter access to compromise hypervisors and control the entire virtual infrastructure domain.

CVE-2026-55040 — Microsoft / SharePoint

• What it is: A weak authentication implementation allowing remote attackers to bypass internal access controls over the network.

• Affected versions: Supported Microsoft SharePoint Server installations (see Microsoft Security Update Guide).

• Exploitation status: Active in the wild (CISA KEV catalog).

• Patch available: Yes — vendor security update.

• CISA due date: September 1, 2026.

• Operational risk: Threatens internal document repositories. Allows unauthorized network actors to bypass security features and access restricted tenant metadata or internal sites.

CVE-2026-65400 — Apple / macOS

• What it is: An improper authentication handling flaw that allows unauthenticated local network actors to access macOS Screen Sharing sessions.

• Affected versions: macOS versions prior to security update patches.

• Exploitation status: Active in the wild (CISA KEV catalog).

• Patch available: Yes — macOS platform security updates.

• CISA due date: September 1, 2026.

• Operational risk: High risk in enterprise desktop networks and hybrid environments where local network segmentation is missing or broad.

CVE-2025-62593 — Ray-Project / Ray

• What it is: A code injection flaw enabling unauthenticated remote attackers to execute arbitrary commands inside Ray execution nodes.

• Affected versions: Ray-Project Ray installations lacking API authentication.

• Exploitation status: Active in the wild (CISA KEV catalog).

• Patch available: Partial / Configuration dependent — upgrade software and enforce dashboard authentication.

• CISA due date: August 31, 2026.

• Operational risk: Exposes high-performance compute clusters and model training environments directly to complete cluster takeovers.

Exploitation Context

Confirmed threat intelligence reporting indicates active, targeted campaigns targeting enterprise collaboration platforms and remote management surfaces. Specifically, security researchers at Kaspersky identified the threat cluster Head Mare actively exploiting TrueConf Server flaws (CVE-2026-72529 and CVE-2026-72530) in the wild to drop custom implants, including the PhantomCore remote access tool.

Additionally, telemetry confirms threat actors are actively scanning perimeter assets for unauthenticated IKE endpoints (CVE-2026-33824) and exposed Zimbra SMTP configurations (CVE-2026-73570) to establish initial footholds. Once access is established, adversaries are targeting unauthenticated internal AI nodes (MLflow and Ray) to query cloud metadata services for privilege escalation.

Remediation Priorities

Detection and Monitoring Guidance

• CVE-2026-73570 (Zimbra):

  • Log Sources: Mail transfer agent (MTA) logs, system audit logs (/var/log/zimbra.log).

  • Behavioral Indicators: Look for unexpected process execution (e.g., /bin/sh, curl, wget) spawned as child processes of the Zimbra SNMP daemon or mail process.

  • Monitoring Gap: Standard endpoint security rarely monitors internal containerized mail processes; deploy auditd host-level execution tracing.

• CVE-2026-72530 & CVE-2026-72529 (TrueConf):

  • Log Sources: Network firewalls, TrueConf Server application event logs, host process creation events.

  • Behavioral Indicators: Detect incoming traffic on TCP port 4307 originating from external or non-administrative subnets. Monitor for process trees where TrueConf services execute cmd.exe, powershell.exe, or file-writing scripting engines.

  • Detection Artifacts: Look for persistence indicators associated with the PhantomCore payload family.

• CVE-2026-64849 (MLflow):

  • Log Sources: Web application firewall (WAF) logs, HTTP access logs, cloud VPC flow logs.

  • Behavioral Indicators: Monitor HTTP POST requests reaching /api/2.0/mlflow/webhooks/*/test. Trace subsequent outbound requests from the MLflow host targeting 169.254.169.254 or internal RFC 1918 subnets.

• CVE-2026-33824 (Microsoft IKE):

  • Log Sources: Windows System Event Log, IPsec / IKE service logs.

  • Behavioral Indicators: Look for Event ID crashes or unexpected restarts of the IKEEXT service. Track malformed or unusually fragmented inbound UDP packets on ports 500 and 4500.

Hackerstorm Analysis

This week’s additions highlight an aggressive expansion into developer and operational AI pipelines. Defenders often hyper-focus on traditional perimeter operating systems while leaving machine learning management frameworks (MLflow, Ray) exposed on internal subnets without authentication. Attackers are exploiting this operational bias—using unauthenticated perimeter access points to enter the network, then instantly leveraging internal SSRF and code injection flaws in AI infrastructure to capture cloud IAM roles. Security teams must recognize that AI platforms are high-value operational targets and apply strict network containment equal to core management controllers.

What you should do next

Ensure you understand the latest KEV additions to cover off any potential gaps in your remediation activities.  View all weekly reports here

Further Reading

The following resources offer technical analysis to help integrate CISA KEV data into operational triage workflows and stay aligned with updated mitigation guidance.

About This Report

Attribution Note

This analysis is based on publicly available reporting and security research summaries. Some technical details may change as additional information becomes available. 

Author Information

Timur Mehmet | Founder & Lead Editor

Timur is a veteran Information Security professional with a career spanning over three decades. Since the 1990s, he has led security initiatives across high-stakes sectors, including Finance, Telecommunications, Media, and Energy. Professional qualifications over the years have included CISSP, ISO27000 Auditor, ITIL and technologies such as Networking, Operating Systems, PKI, Firewalls. For more information including independent citations and credentials, visit our About page.

Contact: This email address is being protected from spambots. You need JavaScript enabled to view it.

Editorial Standards

This article adheres to Hackerstorm.com's commitment to accuracy, independence, and transparency:

• Fact-Checking: All statistics and claims are verified against primary sources and authoritative reports
• Source Transparency: Original research sources and citations are provided in the References section below
• No Conflicts of Interest: This analysis is independent and not sponsored by any vendor or organization
• Corrections Policy: We correct errors promptly and transparently. Report inaccuracies to This email address is being protected from spambots. You need JavaScript enabled to view it.

Editorial Policy: Ethics, Non-Bias, Fact Checking and Corrections

Learn More: About Hackerstorm.com | FAQs

Source Transparency

• CISA Known Exploited Vulnerabilities Catalog: https://www.cisa.gov/known-exploited-vulnerabilities-catalog

• NVD Entry for CVE-2026-73570: https://nvd.nist.gov/vuln/detail/CVE-2026-73570

• NVD Entry for CVE-2026-72530: https://nvd.nist.gov/vuln/detail/CVE-2026-72530

• NVD Entry for CVE-2026-64849: https://nvd.nist.gov/vuln/detail/CVE-2026-64849

• NVD Entry for CVE-2026-33824: https://nvd.nist.gov/vuln/detail/CVE-2026-33824

• Kaspersky / Securelist Threat Analysis (Head Mare Campaign): TrueConf Server exploitation context.

Analyst Notes

• No ransomware attribution was publicly associated with this week's KEV additions at the time of publication.
• No confirmed threat actor attribution was available beyond CISA's confirmation of active exploitation for the majority of entries.
• Where vendor advisory details or EPSS values were unavailable at publication time, placeholders were retained pending analyst validation and vendor updates prior to final publication.