FreeScout CVE-2026-41905

Exploitation requires an authenticated FreeScout user account (PR:L in CVSS vector), meaning the attacker must possess valid login credentials for any user role within the system - not necessarily administrator privileges. The vulnerable Helper::sanitizeRemoteUrl() function must be invoked through a user-accessible feature that accepts URL parameters, such as remote attachment fetching, avatar URLs, webhook endpoints, or external integration configurations. The FreeScout server must have outbound HTTP network connectivity to both the attacker's redirect server and the target internal resources. Exploitation is NOT limited by user interaction (UI:N) or complex preconditions (AC:L), making it straightforward once authentication is achieved. Exploitation effectiveness is highest when FreeScout is deployed in cloud environments with accessible metadata services (AWS EC2, Azure IMDS, GCP metadata) or in networks containing sensitive internal HTTP services on RFC1918 addresses. Default FreeScout installations are vulnerable without configuration changes. Network-level restrictions blocking outbound access to RFC1918 ranges and 169.254.169.254 would prevent exploitation against those specific targets but not against other internal services on routable IPs.

The CVSS 3.1 score of 7.7 (High) reflects significant real-world risk with the vector AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:N/A:N. Network attack vector (AV:N) means exploitation over standard network protocols without physical or local access. Low attack complexity (AC:L) indicates no special conditions beyond authentication. Low privileges required (PR:L) means any authenticated FreeScout user can exploit this - not just administrators. Changed scope (S:C) is critical here: the vulnerability allows escaping FreeScout's security boundary to access resources beyond its normal authorization domain (internal networks, metadata services). High confidentiality impact (C:H) reflects potential access to cloud credentials, internal API keys, database credentials, and sensitive internal service data. The zero integrity and availability impact indicates this is purely an information disclosure attack. In cloud environments (AWS, Azure, GCP), SSRF to metadata services commonly yields IAM credentials enabling full account compromise. The PR:L requirement significantly raises the bar versus unauthenticated SSRF, limiting exploitation to scenarios where attackers have valid user accounts (via phishing, credential stuffing, or insider threat). No CISA KEV listing and no identified public exploit code suggest exploitation is not yet widespread, but SSRF techniques are well-documented and exploitation barriers are low once authentication is achieved. Organizations running FreeScout in cloud environments or with access to sensitive internal networks face elevated risk.

An authenticated FreeScout user (potentially a low-privilege support agent or an attacker who obtained credentials via phishing) exploits the vulnerability by supplying a URL to a feature that invokes Helper::sanitizeRemoteUrl() (likely file attachment fetching, webhook configuration, or integration settings). The attacker provides a URL to their controlled redirect service, such as http://attacker.com/redirect, which initially passes FreeScout's host validation. The redirect service responds with HTTP 302 redirecting to http://169.254.169.254/latest/meta-data/iam/security-credentials/FreeScoutRole. FreeScout's cURL follows the redirect and fetches AWS IAM temporary credentials, which are returned to the attacker. The attacker then uses these credentials to access the organization's AWS resources, potentially gaining full control of cloud infrastructure. Alternatively, the attacker redirects to internal services like http://192.168.1.100:8080/admin/database/backup to exfiltrate database backups or http://10.0.0.5:6379/ to probe Redis instances.