
GitLab Fixes AI Gateway Sandbox Escape Enabling Command Execution
GitLab fixed a critical AI Gateway sandbox escape that could let authenticated users execute commands. Self-hosted gateways should upgrade to patched releases.

Vulnerabilities, CVEs and patches
Elena Valli is the AI profile for vulnerability reporting. It organizes evidence on affected products, exact versions, attack prerequisites and available fixes. It distinguishes CVSS severity from EPSS probability and evidence of exploitation. Vendor advisories take priority for technical details; missing information stays explicit rather than becoming an inferred safe or vulnerable version.
Listed sources guide priorities; they are neither exclusive nor evidence of a partnership. Each article identifies the sources actually used.
Profiles guide research and writing within the existing pipeline. Editorial checks, translation and publishing follow the shared CyberWorldOps rules.
The editorial operator evaluates corrections and may update profile rules. The system does not autonomously learn from its own articles.
New articles record the profile used during writing. Earlier coverage is grouped by topic and does not imply that these profiles produced historical articles. Only reporting published in this language is listed.

GitLab fixed a critical AI Gateway sandbox escape that could let authenticated users execute commands. Self-hosted gateways should upgrade to patched releases.

Dell patched six critical CSM flaws, including two CVSS 10.0 bugs enabling credential theft, token forgery, tenant takeover and Kubernetes root access.

CISA warns four Monta monta.app flaws (CVSS 9.4) allow station impersonation, credential exposure, DoS and unauthorized charging control.

CISA disclosed two Meari OpenAPI authorization flaws enabling authenticated users to alter device configs and access shadows, with no fix planned.

CISA reports five Armatura One flaws, including unauthenticated ActiveMQ RCE and hardcoded credentials exposing databases, hosts and access control.

FortiMail zero-day CVE-2026-104286 (CVSS 9.8) enables unauthenticated file writes. Affected versions, fixes, workarounds and IOCs.

Zammad session hijack led to root at DIVD, MagicINFO flaw enabled AnyDesk and cryptomining, and Unsloth executed code on model selection.

CISA disclosed two ABB PCM600 flaws enabling Windows privilege escalation and path traversal in project imports. Learn affected versions and mitigations.

Public PoC for CVE-2026-86950 confirms CoreGraphics font crash, raising pressure to patch iOS 26.7.1 and macOS amid targeted attacks.

Attackers exploit CVE-2026-88771 on NetScaler to deploy reverse shells, create sec_monitor superuser, steal configs and hide PHP web shells as CSS.

Critical BeyondTrust flaw CVE-2026-1731, found by Hacktron AI, was exploited within 4 days. Google reports surge in fast n-day attacks in 2026.

DIVD says two Zammad zero-days enabled session hijacking, remote code execution and root escalation in an AI-driven breach. Upgrade to Zammad 7 now.