<- ALL CYBER NEWS
Critical
Citrix, NetScaler, ADC, Gateway, Authentication Bypass, Edge, VPN

NetScaler CVE-2026-19490: Auth Bypass, Patch Now
NetScaler CVE-2026-19490 (CVSS 9.3): an unauthenticated auth bypass on Citrix ADC/Gateway, with DoS CVE-2026-19489. Affected builds, fixes, exposure checks.
NetScaler CVE-2026-19490 is a CVSS 9.3 unauthenticated authentication-bypass flaw in Citrix NetScaler ADC and Gateway that lets a remote attacker slip past login on appliances configured as a Gateway or AAA virtual server, and Citrix is urging immediate patching. It was disclosed alongside CVE-2026-19489, a CVSS 8.8 memory-overflow denial-of-service bug. There is no supported workaround for either; upgrading to the fixed firmware is the only remediation, and Citrix-managed cloud instances were patched automatically.
The higher-severity flaw, CVE-2026-19490 (CWE-288, authentication bypass by alternate path), affects appliances acting as a Gateway (SSL VPN, ICA Proxy, CVPN, RDP Proxy) or an AAA virtual server. On recent firmware branches the vulnerable condition specifically requires a SAML Action to be configured, while older builds can be exposed by having a Gateway or AAA vServer at all. The companion bug, CVE-2026-19489 (CWE-119, memory overflow), is a denial-of-service issue exploitable only when SIP ALG is enabled within Large Scale NAT groups. Both affect ADC and Gateway on 14.1 before 73.32 and 13.1 before 63.21.
NetScaler edge appliances have a long, painful exploitation history, and the community reaction reflected it: comparisons to CitrixBleed (CVE-2023-4966) and the 2019 "Shitrix" flaw were immediate. The important distinction is that CitrixBleed leaked live session tokens from memory, so patching alone did not evict attackers who had already stolen sessions. CVE-2026-19490 is a direct bypass of the authentication logic rather than a token leak, so the patch fixes the flaw itself, though killing active sessions after upgrade remains sound hygiene. With tens of thousands of NetScaler devices exposed to the internet and auth-bypass logic bugs being comparatively easy to reverse-engineer from patch diffs, defenders are treating this as an emergency-maintenance event.
Which NetScaler versions are affected, and how do I check exposure?
Upgrade ADC and Gateway to a fixed build; there is no policy or WAF workaround Citrix will support. To scope exposure, inspect your running configuration for the vulnerable conditions before you schedule the window.
Item | Detail |
|---|---|
CVE-2026-19490 | CVSS 9.3, CWE-288 auth bypass (Gateway / AAA vServer; SAML Action on recent builds) |
CVE-2026-19489 | CVSS 8.8, CWE-119 memory-overflow DoS (SIP ALG in LSN groups) |
Affected | ADC and Gateway: 14.1 before 73.32, 13.1 before 63.21 |
Fixed builds | 14.1-73.32, 13.1-63.21 (plus FIPS/NDcPP equivalents) |
Workaround | None; firmware upgrade only |
To check the auth-bypass exposure, review the config for Gateway/AAA virtual servers and SAML actions (for example show run | grep -iE "authentication samlAction|authentication vserver|vpn vserver"), and for the DoS bug check for SIP ALG on LSN groups. NetScaler Console (ADM) flags vulnerable instances automatically.
Should I rotate credentials and kill sessions after patching?
Patching CVE-2026-19490 fixes the authentication logic directly, unlike CitrixBleed which required explicit session termination to be safe. Even so, if the appliance was internet-exposed, incident-response best practice is to terminate active sessions after the upgrade (kill aaa session -all, kill icaconnection -all) and, if you see any sign of unauthorized access, rotate SAML signing certificates, IdP keys, and any service-account credentials bound to the AAA vServers. Hunt ns.log and HTTP error logs for spikes in unauthenticated requests to SAML/AAA endpoints and for unexpected authentication successes.
Our read
Perimeter appliances remain the softest hard target: a single internet-facing box that terminates VPN and SSO for the whole workforce, patched on a change-control cadence that attackers do not share. The recurring lesson from Citrix, and the reason continuous verification exists as a pillar, is that "we applied the firmware" is a claim, not evidence. Prove the vulnerable config conditions are gone, prove no unauthenticated request reached the SAML path during the exposure window, and prove sessions were cycled. On an edge device that fronts your entire identity flow, the difference between assuming and verifying is the difference between a maintenance window and an incident.
Reporting by BleepingComputer and The Hacker News; CVSS and affected-version data per Citrix's advisory and NVD. Sources linked above.
Related: Cisco ASA/FTD CVE-2026-20349 exploited and SharePoint CVE-2026-50522 active exploitation.