Cisco Discloses Its Seventh SD-WAN Zero-Day of 2026, with No Patch in Sight
Cybersecurity

Cisco Discloses Its Seventh SD-WAN Zero-Day of 2026, with No Patch in Sight

Cisco has disclosed CVE-2026-20245, a command-injection zero-day in Catalyst SD-WAN Manager that grants root access and is already being exploited in the wild, with no patch available. It is the seventh SD-WAN zero-day Cisco has disclosed this year, and the fourth flagged as actively exploited by CISA.

PublishedJune 7, 2026
Read time6 min read
Share

Seven Zero-Days in One Product in One Year

The cadence of Cisco SD-WAN zero-day disclosures in 2026 has moved from alarming to something approaching a pattern that demands a structural explanation. CVE-2026-20245, disclosed in June and reported as actively exploited by Google Cloud's Mandiant team, is the seventh such vulnerability to emerge from Cisco's Catalyst SD-WAN Manager this year. It follows CVE-2026-20182, an authentication bypass; CVE-2026-20133, an information disclosure flaw; CVE-2026-20127, an authentication bypass traced back to 2023; and three additional disclosures. CISA has now flagged four of the seven as actively exploited.

The specific mechanics of CVE-2026-20245 follow a familiar template. The vulnerability lives in insufficient input validation in the SD-WAN Manager command-line interface. An attacker with netadmin-level access can upload a crafted file and trigger command injection, escalating to root on the affected system. Cisco's own advisory summarises the impact with economy: "A successful exploit could allow the attacker to perform command injection attacks on an affected system and elevate their privileges as the root user." At root, an attacker can push configuration changes to any of the edge devices the manager controls — a network of up to 6,000 devices in large enterprise deployments.

Why Netadmin Access Is Not a Meaningful Barrier

Cisco's disclosure notes that exploitation requires netadmin privileges, which might read as a meaningful access barrier. It is not. Netadmin credentials can be obtained in at least three ways that are well within the capability of the threat actors currently targeting enterprise SD-WAN infrastructure. First, credential theft through phishing or the Miasma-class supply chain attacks documented elsewhere this week. Second, exploitation of the authentication bypass vulnerabilities CVE-2026-20182 or CVE-2026-20127 — both of which are already publicly disclosed and, in the case of CVE-2026-20127, have been in the wild since 2023. Third, purchase on initial access broker markets, where SD-WAN manager credentials for named enterprise targets are regularly listed.

The observed attacks have confirmed the escalation path in production. Cisco noted limited cases where exploitation resulted in configuration changes being pushed to edge devices — meaning the attacker successfully reached root on the manager and then used that access to modify the network behaviour of downstream devices. In a branch-heavy enterprise with hundreds or thousands of SD-WAN edges, configuration-level control of the manager is effectively control of the entire WAN fabric, including traffic steering, policy enforcement, and security posture.

Mandiant's Discovery and What It Implies

The attribution of discovery to Mandiant — now a Google Cloud subsidiary — is significant context. Mandiant's SD-WAN threat intelligence practice has been tracking a cluster of activity targeting enterprise networking infrastructure throughout 2026. Their involvement in this disclosure suggests the vulnerability was found during an active incident response engagement rather than through routine research, which typically indicates a more mature attacker with real operational objectives rather than proof-of-concept exploitation.

We do not yet have public attribution for the threat actor or actors behind the observed CVE-2026-20245 exploitation. The pattern of targeting SD-WAN manager infrastructure — which by design sits at the control plane of enterprise network operations — is consistent with nation-state-affiliated espionage groups that have historically prioritised network device compromise as a persistence mechanism. Once present at root on a network manager, an actor can observe traffic flows, intercept credentials, and maintain access that survives endpoint remediation campaigns.

No Patch, No Workaround — and a Remediation Problem

As of the date of this disclosure, Cisco has not released a security patch for CVE-2026-20245. The company's guidance centres on three actions: harden access controls around the SD-WAN manager CLI, preserve forensic evidence of any anomalous activity, and conduct a compromise review to determine whether exploitation has already occurred in your environment. The absence of a workaround — unlike some Cisco advisories which offer feature disablement as a temporary mitigation — means that any organisation running Catalyst SD-WAN Manager is carrying live exposure until a patch ships.

The timing of patch availability is unknown. Given that this is the seventh zero-day in the product this year and that Cisco's security response cadence has already been strained by the preceding disclosures, security teams should not assume a patch will arrive quickly. Network segmentation of SD-WAN manager infrastructure, strict IP allowlisting for CLI access, mandatory multi-factor authentication for any account with netadmin privileges, and continuous monitoring for unexpected configuration pushes to edge devices are the operational controls available in the absence of a vendor fix.

The SD-WAN Attack Surface in Context

Cisco SD-WAN Manager is not alone in attracting this level of attention from threat actors in 2026. Ivanti, Palo Alto Networks, and Fortinet have all disclosed critical vulnerabilities in network management and edge infrastructure products this year. The pattern across vendors is consistent with a deliberate adversarial focus: network management planes are high-value targets because they provide leverage over an entire organisation's connectivity rather than a single endpoint. Compromise at the management layer is qualitatively different from endpoint compromise.

What distinguishes the Cisco SD-WAN situation is the serial nature of the disclosures. Seven zero-days in one product in one year suggests either a systematic adversarial research campaign against that specific product, or a structural code quality problem that is being discovered incrementally. Either hypothesis has significant implications. In the first case, sophisticated actors have decided that Cisco SD-WAN Manager is worth sustained research investment. In the second, the remediation effort required goes beyond individual patches and into architectural review. Cisco has not yet commented publicly on which hypothesis better fits the evidence.

Immediate Actions for Security and Network Teams

For organisations running Cisco Catalyst SD-WAN, the response priority is assessment before remediation. First, determine whether any account with netadmin privileges has shown unexpected authentication events, CLI sessions, or file upload activity in the past 30 days. Second, review edge device configuration logs for changes that cannot be traced to an authorised change ticket. Third, isolate SD-WAN manager access to a jump host with MFA and IP restrictions if that is not already in place.

We are tracking this story as it develops. Given the pace of SD-WAN zero-day disclosures in 2026, organisations with significant Cisco SD-WAN deployments should establish a standing response protocol for rapid assessment when new CVEs are disclosed, rather than treating each disclosure as a one-off event. The seventh zero-day of the year will not be the last.

Tagged#news#security#cisco#sd-wan#zero-day#cve#vulnerabilities#enterprise