Residential gateways and broadband routers form the first and often only line of defense protecting home and small office networks from direct Internet exposure. Security researchers have disclosed an unpatched vulnerability in widely deployed Calix GS7 XGS (GS5239XG) residential gateways that allows unauthenticated remote attackers on the public Internet to bypass Network Address Translation (NAT) and expose internal devices.

Universal Plug and Play exposed to the public Internet

The vulnerability, tracked under CVE-2026-75501 (and CERT/CC vulnerability note VU#756733), stems from a fundamental implementation flaw in the router’s Universal Plug and Play (UPnP) service within firmware version EXOS/6.6.47.

UPnP was originally designed for local area networks (LANs) to allow trusted internal devices – such as gaming consoles, media servers, and smart televisions – to automatically request dynamic port forwarding from the gateway without requiring manual user configuration. To ensure security, UPnP daemon services should only listen on internal LAN interfaces and strictly reject requests originating from the wide area network (WAN) or public Internet.

In affected Calix GS7 XGS units, the UPnP WANIPConnection service is mistakenly bound to the external WAN interface and lacks authentication. As a result, an attacker anywhere on the Internet can send standard UPnP SOAP control messages directly to the router’s public IP address.

How attackers abuse remote port mapping

By issuing unauthenticated AddPortMapping requests to the exposed UPnP service, remote threat actors can arbitrarily manipulate the router’s NAT routing tables. An attacker can map any external port on the public WAN IP to any internal IP address and port on the local private network (e.g., 192.168.1.0/24).

This allows attackers to punch holes through the router’s firewall and directly expose vulnerable local devices to the Internet, including:

The mechanics of the UPnP SOAP exploit

The technical exploit requires only a simple HTTP POST request directed at the router’s WAN IP on the UPnP listening port. The payload contains a standard XML body invoking the AddPortMapping action on the urn:schemas-upnp-org:service:WANIPConnection:1 namespace.

Because the gateway fails to validate whether the source IP belongs to an internal RFC 1918 subnet, the router processes the request, commits the new port-forwarding rule into its active iptables NAT chains, and immediately begins forwarding inbound external traffic to the specified internal target host. The entire transaction takes milliseconds and generates no security alerts on the local network.

The broader risk to remote workforces and enterprise VPNs

The shift to hybrid work has turned home networks into extensions of corporate perimeters. When a residential gateway is compromised or punctured via NAT bypass, an attacker gains direct visibility into internal devices on the same subnet as an employee’s work laptop.

If an enterprise laptop connects to the corporate network via split-tunnel VPN, an attacker who forwarded ports to local services (such as unpatched Windows SMB, RPC, or developer servers running on port 3000 or 8080) can initiate direct exploitation attempts against the workstation. From there, malicious actors can harvest session tokens, deploy ransomware, or establish persistence across the broader enterprise environment.

Widespread deployment across broadband providers

Calix residential gateways are widely supplied by fiber-to-the-home (FTTH) broadband service providers across North America and globally. Because these devices are typically managed directly by Internet Service Providers (ISPs) via TR-069 or TR-369 protocols, residential end users often lack direct administrative control over firmware updates or low-level UPnP daemon settings.

Public proof-of-concept exploit scripts demonstrating remote NAT table manipulation have been released, significantly raising the risk of automated scanning and opportunistic exploitation across ISP subscriber ranges.

Analyzing the threat surface and automated scanning activity

Because UPnP runs over standard HTTP protocols, vulnerability scanners and botnet operators can trivially identify exposed Calix endpoints. A simple HTTP GET request to the root device descriptor endpoint reveals the device model, firmware build, and service control URLs.

Once identified, threat actors can integrate these devices into residential proxy networks or weaponize them for distributed denial of service (DDoS) amplification. The ability to expose arbitrary internal ports transforms trusted residential gateways into unmonitored ingress tunnels, presenting severe risks for managed service providers and enterprise SOC teams defending hybrid work environments.

Mitigation steps for network administrators and ISPs

Until Calix and upstream broadband carriers deploy patched EXOS firmware updates across subscriber fleets, network administrators and affected users should implement defensive mitigations:

Leave a Reply

Your email address will not be published. Required fields are marked *