GlobalProtect App CVE-2026-0297

Exploitation requires the attacker to be positioned as a man-in-the-middle between the GlobalProtect client and a gateway, or to operate a rogue GlobalProtect gateway reachable by the target client - for example, via ARP spoofing, DNS poisoning, rogue Wi-Fi access point, or compromise of VPN infrastructure. This positional prerequisite (encoded as AV:A and AT:P in the CVSS 4.0 vector) is the primary limiting factor: mass internet-based exploitation is not possible. The GlobalProtect app must actively initiate or process a connection to the attacker-controlled endpoint for the overflow to trigger; passive bystander clients not currently connecting are not at risk in that moment. No authentication is required on the attacker side (PR:N), and no user interaction beyond the routine VPN connection is needed (UI:N). Exploitation is limited to systems running the GlobalProtect app on Windows, macOS, or Linux; specific version bounds must be verified via the vendor advisory. Systems not using GlobalProtect or isolated from adjacent-network threat actors are not exposed.

The CVSS 4.0 base score of 5.2 with vector AV:A/AC:L/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N reflects a dual-signal tension: impact is maximally severe (full system compromise across all CIA pillars) but the attack surface is meaningfully constrained by adjacency and positional requirements. AV:A limits exploitation to attackers already on the same network segment or capable of intercepting gateway-bound traffic; AT:P further requires active MitM positioning or rogue gateway infrastructure, neither of which is trivially automated. The supplemental E:U (Exploit maturity: Unreported) and AU:N (not automatable) confirm no mass-exploitation or weaponized tooling has surfaced. KEV status is absent from available data, meaning CISA has not confirmed in-the-wild exploitation. The U:Amber urgency rating from Palo Alto signals a measured but definite response requirement - not an emergency patch sprint, but not deferrable. EPSS data is not provided, so probabilistic exploitation likelihood cannot be independently assessed. For organizations with distributed remote workforces relying on GlobalProtect for corporate network access, a compromised VPN client operating at root/SYSTEM represents a full endpoint takeover vector - the constrained attack prerequisites do not diminish the impact if they can be met.

An attacker on the same network segment as a corporate endpoint user performs ARP spoofing or DNS hijacking to redirect GlobalProtect VPN traffic toward an attacker-controlled server mimicking a legitimate gateway. When the GlobalProtect app initiates a connection and processes the crafted gateway response containing an oversized or malformed payload, the out-of-bounds write corrupts process memory, enabling the attacker to redirect execution flow and spawn a shell with SYSTEM privileges on Windows or root on macOS and Linux. No public proof-of-concept code has been identified at time of analysis.