Loki Intelligence — Security Briefs · Published
Daily Brief: FortiMail, Zammad, SD-WAN and GitLab AI Gateway Flaws
Exploitation is spreading across mail security gateways, helpdesk software and network management platforms, while a critical GitLab AI Gateway flaw shows that AI workflow tooling now carries server-level risk. Teams should prioritize FortiMail, Zammad and SD-WAN Manager remediation, update self-hosted AI gateways, and report exposure to leadership in business terms rather than activity counts.
Signal 01 · CISA
CISA Adds Two Known Exploited Vulnerabilities to Catalog
CISA added two Zammad vulnerabilities to the KEV catalog based on evidence of active exploitation: CVE-2026-102489, a session fixation flaw, and CVE-2026-102490, an improper privilege management flaw. The catalog entries note the pair can be chained, moving an attacker from code execution as the zammad user to root.
Why it matters: Teams running self-hosted Zammad should apply vendor fixes immediately, limit internet exposure, and review hosts for unexpected sessions or privilege changes, since helpdesk tickets often hold sensitive customer data.
Source: CISA
Signal 02 · The Hacker News
CISA Adds Exploited Cisco Catalyst SD-WAN Manager Auth Bypass to KEV
The Cisco Catalyst SD-WAN Manager authentication bypass CVE-2026-76504 is now on CISA's KEV list, with federal agencies given until October 3 to fix it. watchTowr noted that eight Cisco SD-WAN CVEs have reached KEV in 2026, a sign of how attractive centralized network management consoles are to attackers.
Why it matters: Repeated exploitation of one platform means SD-WAN Manager needs a standing fast-patch process, strict management-plane access controls, and routine log review for unauthorized logins.
Source: The Hacker News
Signal 03 · The Hacker News
GitLab Patches Critical 9.9 AI Gateway Flaw Allowing Command Execution on Self-Hosted Servers
GitLab fixed CVE-2026-90970, a CVSS 9.9 flaw in its AI Gateway that could let an authenticated user with Duo Agent Platform access escape the custom-flow prompt template sandbox and run commands on the gateway. Only organizations hosting their own gateway must act; fixed versions are 19.2.4, 19.3.2 and 19.4.1.
Why it matters: Update self-hosted AI Gateway images, review who holds Duo Agent Platform access, and treat the gateway's JWT signing keys and model-provider connections as sensitive assets.
Source: The Hacker News
Signal 04 · BleepingComputer
GitLab warns of critical RCE vulnerability in AI Gateway service
GitLab urged customers running a self-hosted AI Gateway to patch a critical improper neutralization flaw that can lead to arbitrary command execution by users with basic privileges and Duo Agent Platform access. Customers relying on GitLab-hosted gateways are already protected, and GitLab contacted affected self-hosters before disclosure.
Why it matters: AI integration services often sit outside normal patch inventories; confirm whether a self-hosted gateway exists, upgrade it to a fixed version, and limit who can create agent flows.
Source: BleepingComputer
Signal 05 · CISA
CISA Adds One Known Exploited Vulnerability to Catalog
CISA's October 1 KEV update covers CVE-2026-104286, a Fortinet FortiMail path traversal vulnerability with evidence of active exploitation. CISA encourages all organizations, not only federal agencies, to adopt risk-based vulnerability management and prioritize remediation of KEV-listed flaws.
Why it matters: Mail gateways are exposed by design, so inventory FortiMail instances, apply fixes or vendor workarounds, and check for unexpected files before declaring appliances clean.
Source: CISA
Signal 06 · The Hacker News
Why CISOs Struggle to Answer the Board's Three Hardest Questions, and How to Fix the Report
This sponsored piece argues that board reporting built on activity counts cannot answer how secure the organization is, what its financial exposure is, or whether posture is improving. It recommends defining critical assets with the business, correlating data from existing tools, mapping attack paths, and reporting exposure trends in financial terms.
Why it matters: Security leaders can make reporting more useful by tying remediation to reachable critical assets, including non-human and AI identities, rather than to raw finding counts.
Source: The Hacker News
Signal 07 · The Hacker News
Critical FortiMail Zero-Day Flaw Exploited in Attacks Allows Unauthenticated Arbitrary File Writes
Fortinet confirmed in-the-wild exploitation of CVE-2026-104286, a CVSS 9.8 FortiMail flaw that lets unauthenticated attackers write arbitrary files through crafted web requests. Several fixed releases were still upcoming, so Fortinet advises disabling the IBE feature and restricting management interface access in the meantime, and has shared indicators of compromise.
Why it matters: Apply available updates or the vendor workarounds now, block internet access to FortiMail administration, and compare appliances against Fortinet's published IP and file indicators.
Source: The Hacker News
Brief sources
Related briefs
- Hardware Hacking for Embedded Security
- How to Evaluate an AI Security Testing Provider
- What Is AI Agent Security Testing?
Relevant Loki services: Orvyn — AI agent security private preview and SKYEN web & API pentesting.