Novel 'TunnelVision' Attack Bypasses VPN Encryption by Exploiting DHCP Protocol Design Flaws
Security researchers have demonstrated a technique that completely circumvents VPN encryption on most operating systems by manipulating DHCP option 121, potentially exposing years of supposedly protected traffic to network-level surveillance.
TunnelVision Attack Exposes Fundamental VPN Vulnerability
Cybersecurity researchers have disclosed a novel attack technique that effectively neutralizes VPN encryption protections on Windows, Linux, macOS, and iOS devices by exploiting fundamental design characteristics of the Dynamic Host Configuration Protocol. The attack, designated 'TunnelVision,' works by manipulating DHCP option 121 to redirect traffic outside the encrypted VPN tunnel without user awareness.
The technique is deceptively simple yet devastatingly effective. When a device connects to a network, it receives configuration information via DHCP, including routing tables that determine how traffic flows. By sending malicious DHCP option 121 routes, an attacker controlling the local network can instruct the device to send traffic directly to the attacker rather than through the VPN tunnel—all while the VPN interface reports a secure, encrypted connection.
What makes TunnelVision particularly dangerous is its position below the VPN software layer. Because the attack manipulates the operating system's routing table directly, VPN applications have no visibility into the compromise. Users see a connected, encrypted VPN tunnel while their actual traffic flows in cleartext across the local network.
The vulnerability affects virtually all commercial and open-source VPN implementations, including WireGuard, OpenVPN, and proprietary enterprise solutions. Android remains the only major operating system immune to the attack due to its network namespace isolation architecture.
Practical exploitation requires attacker control of the local network's DHCP server—typical in hostile Wi-Fi scenarios like airports, hotels, or cafes, but also achievable in compromised corporate environments. Several VPN providers have begun developing mitigations, including DHCP option filtering and routing table monitoring, though comprehensive protection requires operating system-level changes.
The disclosure raises uncomfortable questions about VPN trust models that have dominated security advice for the past decade. For threat models assuming hostile network infrastructure—journalists, activists, or corporate travelers—the attack represents a fundamental breakdown in assumed protections.
Until widespread mitigations are deployed, security-conscious users should consider cellular connections or trusted networks as the only reliable protection layer beneath VPN encryption.