This week’s intelligence highlights severe escalations across core web dependencies, perimeter firewalls, and cloud orchestration platforms. Threat actors are weaponizing fundamental compression libraries for remote code execution, deploying autonomous AI agents to automate ransomware deployment, and utilizing ephemeral multi-hop routing to breach enterprise control planes.

SISA Weekly Threat Watch – our weekly feature brings to you a quick snapshot of all the major security vulnerabilities that posed a threat to organizations worldwide. These recurring actionable threat advisories will also provide information and recommendations that will help security teams take appropriate actions to defend against the latest and critical threats.

1. Critical Edge, Virtualization & Infrastructure Exploits

Perimeter gateways, mesh networks, and hypervisor environments are facing pre-authentication zero-days and active exploitation campaigns designed to grant unauthenticated root access.

• Canonical LXD & Incus CVSS 10.0 Auth Bypass — A critical certificate validation flaw (CVE-2026-112816) allows unauthenticated remote attackers to bypass mutual TLS (mTLS) in Canonical LXD, MicroCloud, and Incus. Attackers can forge self-signed leaf certificates and append trusted intermediary certificates to impersonate administrators, gaining full root access to the underlying host infrastructure.
• Active Cisco ASA/FTD & WatchGuard Exploitation — Threat actors are actively exploiting a zero-day logic flaw (CVE-2026-76418) in the SSL VPN subsystem of Cisco ASA and FTD, forcing immediate device reloads and denial of service. Concurrently, WatchGuard disclosed seven critical pre-authentication flaws in Fireware OS allowing remote code execution via stack-based buffer overflows in WireGuard and Active Directory daemons.
• VMware ESXi Evasion & Tailscale SSH Privilege Escalation — Advanced persistent threats are actively weaponizing an insufficient logging flaw (CVE-2026-41709) in VMware ESXi to manipulate virtual machines and deploy rogue VIBs without leaving audit trails. In zero-trust networks, an audit of Tailscale SSH ("Tailworm") revealed that unprivileged users can exploit socket forwarding race conditions to achieve full root execution.
• PHP2Shell & Fragnesia Firewall Compromise — Attackers are chaining the PHP2Shell vulnerability (CVE-2026-109618/19) with the Fragnesia Linux kernel exploit (CVE-2026-46300) to achieve root-level control over perimeter firewalls running AppWeb.

2. Core Supply Chain & Application RCE

Vulnerabilities in foundational web compression libraries and enterprise integration frameworks are exposing massive segments of the digital supply chain to unauthenticated RCE.

• HTTP2Shell (Brotli Decompression Flaws) — Following the "Node2Shell" disclosure, the CVE-2026-111497 vulnerability in the Google Brotli C-decoder has escalated into an ecosystem-wide crisis ("HTTP2Shell"). Exploitation allows unauthenticated RCE across Node.js, Microsoft .NET, Python, and PHP environments when incoming or outgoing compressed HTTP streams are parsed. Microsoft and Cloudflare have issued emergency out-of-band updates to mitigate the blast radius.
• Apache Camel High-Severity Vulnerabilities — Core components of the Apache Camel enterprise integration framework face critical XXE injection flaws (Camel-Saxon, Camel-Smooks) and unsafe Java object deserialization (Camel-JGroups, CVE-2026-88900). Exploitation permits unauthenticated network attackers to achieve RCE or exfiltrate local system files via SSRF.
• Convertr Payment Supply Chain Compromise — A Brazilian Magecart syndicate compromised the self-hosted analytics infrastructure of Convertr (CVE-2026-8467). Attackers injected a polymorphic digital credit card skimmer into a core tracking script, harvesting payment card data, CVVs, and Brazilian CPF taxpayer IDs across major retail merchants.

3. State-Sponsored Espionage, AI Intrusions & Extortion

Nation-state actors are deploying highly ephemeral attack infrastructure, while financially motivated syndicates utilize autonomous AI agents and advanced identity theft tools.

• Volt Typhoon "Silent Train" & FortiBleed 2.0 — China-nexus group Volt Typhoon has operationalized "Silent Train," an ephemeral multi-hop routing framework utilizing SOHO/IoT devices to bypass geofencing and deploy zero-days against control-plane orchestrators (e.g., Versa Director, Cisco SD-WAN). Concurrently, the "FortiBleed 2.0" dataset exposed 1,703 FortiGate configurations, leaking over 40,000 cleartext secrets including Active Directory LDAP bind credentials.
• GOLD OGRE & NetamLoader — The GOLD OGRE syndicate is deploying Aur0ra ransomware utilizing autonomous AI coding agents (such as Cursor) to execute automated Active Directory reconnaissance and lateral movement. Initial access is often secured via NetamLoader, which abuses .NET AppDomainManager injection and Windows Update directories to execute in-memory payloads and disable AMSI.
• Aokigahara Infostealer & DarkCore RAT — The Aokigahara infostealer specifically targets developer secrets, extracting AWS credentials and SSH keys from AI IDEs (Cursor, Claude). DarkCore RAT bypasses risk-based MFA by utilizing a dedicated session-hijacking browser that tunnels traffic directly through the victim’s IP address.

4. Client-Side Exploitation & CDN Bypasses

• Microsoft Edge Full-Chain Logic Escape — A zero-click exploit chain utilizes five browser logic vulnerabilities in Microsoft Edge to achieve full client-side arbitrary code execution and sandbox breakout without memory corruption, facilitating SSO credential theft and Universal Cross-Site Scripting (UXSS).
• TeamViewer RCE & Cloudflare CSP Bypass — TeamViewer patched critical flaws (CVE-2026-92372) allowing remote endpoints to execute code on viewer machines via malformed display packets. Researchers also exposed a cache-manipulation chain within Cloudflare’s Image Transformations that strips Content Security Policies (CSP) to trigger XSS.

Proactive Steps for the Week

• Remediate Core Runtimes & Compression Middleware: Immediately deploy out-of-band updates for Node.js, Microsoft .NET, and Python environments to patch the HTTP2Shell Brotli decoder flaws. Until patched, configure API gateways and ingress controllers to reject incoming HTTP requests carrying Content-Encoding: br headers.
• Patch Edge Infrastructure & Hypervisors: Apply emergency patches to Canonical LXD/MicroCloud/Incus to prevent the CVSS 10.0 certificate bypass. Upgrade Cisco ASA/FTD devices, WatchGuard Fireware OS, and VMware ESXi hypervisors to their respective fixed releases.
• Rotate Exposed FortiGate & Payment Secrets: Cross-reference organizational domains against the FortiBleed 2.0 exposure dataset and immediately rotate all Active Directory LDAP bind passwords and VPN secrets. E-commerce merchants utilizing Convertr must purge edge CDN and browser caches to eliminate persisted copies of the Magecart skimmer.
• Secure Developer & AI Environments: Restrict the loading of unsigned .config files and custom assemblies to thwart NetamLoader's .NET injection. Ensure SSH keys and cloud SDK profiles are stored in hardware tokens or OS credential vaults to protect against Aokigahara developer targeting.
• Harden Browsers & Remote Access: Verify Microsoft Edge is updated to build 148.0.3967.70 or higher to close the zero-click logic escape. Set TS_SSH_DISABLE_FORWARDING=1 in Tailscale environments to prevent socket forwarding abuse, and update TeamViewer to 15.83+.

Explore our DFIR Solutions to discover how our advanced incident response support, compromise assessments, and threat hunting frameworks can insulate your enterprise infrastructure against these campaigns.