Fortinet FortiSandbox CVE-2025-53679

Exploitation requires the attacker to already hold high-privileged (administrator-level) credentials to the FortiSandbox or FortiSandbox Cloud management interface (CVSS PR:H), and network reachability to its HTTP or HTTPS service from the attacker's position. No user interaction is required (UI:N) and attack complexity is low (AC:L), so once the credential and network preconditions are met, a single crafted HTTP/HTTPS request is sufficient. Limiting factors are therefore the strength and exposure of admin credentials, whether the management interface is restricted to trusthosts or a management network, and whether the deployed version falls within the listed vulnerable ranges (5.0.0-5.0.2, 4.4.0-4.4.7, all 4.2.x, all 4.0.x, Cloud 24.1, all Cloud 23.x).

The CVSS 3.1 score of 7.2 (AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:H) signals that exploitation is network-reachable and low-complexity but requires high privileges, which materially shrinks the threat population to users who already hold administrative credentials on the appliance. There is no EPSS score, KEV listing, or public POC provided in the available intelligence, so this should be treated as a credible but not-yet-actively-exploited issue at time of analysis. The real-world risk is concentrated in scenarios where admin credentials are shared, weakly protected, or already compromised - in those cases the appliance pivots from a detection tool into a full command-execution foothold inside the security stack, which is a high-value outcome for an attacker even though the raw CVSS suggests only High rather than Critical severity.

An attacker who has obtained or coerced FortiSandbox administrator credentials - for example via phishing of a SOC engineer, reuse of leaked credentials, or pivoting from a compromised FortiManager - authenticates to the appliance and issues a crafted HTTP or HTTPS request to a vulnerable admin endpoint whose handler passes a parameter into a shell command. The injected shell metacharacters cause the appliance to execute arbitrary OS commands under the service account, giving the attacker code execution on a trusted security inspection device that sees decrypted samples and intra-environment telemetry. No public POC is available at time of analysis, so exploitation today would require independent reverse engineering of the patched endpoints.