CVE-2025-20333
Critical · 9.9 CISA KEVCVSS
9.9
EPSS
71.2%
Published
Sep 25, 2025
CISA due date
Sep 26, 2025
Cloudanix analysis
Attackers with valid VPN logins can take control of vulnerable Cisco firewall VPN devices, potentially giving them a foothold into the organization’s network.
- Exploitation
- Remote Authenticated
- Attacker needs
- Low Privilege
- Complexity
- Moderate
- Impact
- Full Compromise
- Component
- Network Service
- Class
- RCE
- Remediation
- Upgrade
- Effort
- Medium
- Patch risk
- Medium
What it actually does
CVE-2025-20333 is an authenticated remote code execution vulnerability in the VPN web server component of Cisco Secure Firewall ASA Software and Cisco Secure Firewall FTD Software. The flaw is caused by improper validation of user-controlled input in HTTP(S) requests sent to the VPN service. An attacker who already has valid VPN user credentials can send crafted requests to the exposed VPN web interface and execute arbitrary code with root privileges on the firewall. Successful exploitation can result in full device compromise, including modification of firewall configuration, theft of VPN credentials or certificates, traffic interception, deployment of persistence, and use of the firewall as a pivot point into internal networks.
Why it matters to the business
This is a high-urgency perimeter-device vulnerability because affected ASA and FTD systems commonly sit at the network edge and terminate remote-access VPN sessions. Exploitation gives an attacker control of a security appliance that is trusted to enforce access, inspect traffic, and protect internal systems. Because the vulnerability is listed in CISA KEV, has active exploitation in the wild, and carries a KEV remediation deadline of 2025-09-26, organizations should treat exposed VPN services as potentially targeted. Even though valid VPN credentials are required, stolen, phished, reused, or purchased credentials are common in real-world attacks, making the operational risk significant. Consequences may include unauthorized internal access, outage of VPN or firewall services, data theft, compromise of privileged secrets, and loss of confidence in perimeter controls.
How an attack unfolds
An attacker first obtains valid VPN credentials through phishing, credential theft, password reuse, brokered access, or prior compromise. The attacker then identifies an internet-exposed Cisco ASA or FTD VPN web server running a vulnerable software release. After authenticating to the VPN-facing HTTPS service, the attacker sends specially crafted HTTP(S) requests containing malformed user-supplied input that is not properly validated by the VPN web server. The vulnerable process handles the crafted input in a way that allows attacker-controlled code to run as root on the appliance. With root-level execution, the attacker can alter device configuration, add or modify users, extract VPN session information, collect certificates or keys, disable or evade logging, install persistence, and pivot from the firewall into protected networks. Public reporting and KEV status indicate active exploitation, including activity tracked in Cisco ASA/FTD exploitation campaigns and references to ArcaneDoor-related targeting, so defenders should assume exploitation attempts may be operationalized rather than theoretical.
How to fix it
Upgrade affected Cisco Secure Firewall ASA Software and Cisco Secure Firewall FTD Software to a Cisco-fixed release identified in the official Cisco security advisory or Cisco Software Checker for CVE-2025-20333. No exact fixed versions were provided in the input, so remediation teams should validate the currently installed train and apply the vendor-designated fixed build for that train. Because exploitation is active and the issue is in CISA KEV, prioritize internet-facing VPN appliances and complete remediation immediately, especially where the CISA due date of 2025-09-26 applies. If immediate upgrade is not possible, reduce exposure by disabling unused remote-access VPN or clientless VPN functionality, restricting VPN web access to trusted source networks where feasible, enforcing MFA for all VPN users, disabling inactive accounts, rotating potentially exposed credentials, and monitoring for anomalous authenticated VPN activity. If exploitation is suspected, do not rely on patching alone: preserve evidence, review configuration for unauthorized changes, rotate local and AAA-integrated credentials, renew or replace exposed certificates and keys if necessary, and consider rebuilding or reimaging the appliance from trusted software.
How to confirm the fix worked
Confirm the installed ASA or FTD software version with vendor-supported administrative commands or management tooling and compare it against the Cisco advisory or Cisco Software Checker for this vulnerability. Validate that all exposed VPN-capable appliances are either running a fixed release or do not expose the vulnerable VPN web server functionality. From an external network perspective, verify whether HTTPS VPN portals are reachable and confirm that only intended systems expose them. Review configuration to ensure unused WebVPN, clientless SSL VPN, and remote-access VPN features are disabled where not required. After remediation, confirm successful upgrade, reboot or failover completion where applicable, policy deployment for FTD-managed devices, and restoration of expected VPN functionality. For suspected compromise, validate integrity by comparing running configuration, startup configuration, local users, AAA settings, trustpoints, certificates, scheduled jobs or persistence mechanisms, and logging settings against known-good baselines.
Typical attack path
- Acquire valid VPN user credentials through phishing, credential stuffing, infostealer logs, access brokers, or prior intrusion.
- Scan for internet-facing Cisco ASA or FTD VPN web services over HTTPS.
- Authenticate to the VPN web server using the valid credentials.
- Send crafted HTTP(S) requests to the vulnerable VPN web server component.
- Trigger improper input validation and achieve arbitrary code execution as root on the appliance.
- Establish persistence or modify configuration to maintain access.
- Harvest credentials, certificates, VPN session data, configuration files, or network intelligence from the firewall.
- Use the compromised appliance to pivot into internal systems, intercept traffic, or disrupt security controls.
How to detect it
Investigate both pre-exploitation authentication and post-exploitation device behavior. Collect and review Cisco ASA/FTD syslog, VPN authentication logs, AAA/RADIUS/TACACS/LDAP logs, firewall management logs, FTD management-center events, NetFlow, EDR telemetry from adjacent systems, and upstream web proxy or packet-capture data if available. Look for successful VPN logins from unusual geographies, hosting providers, anonymization services, new user agents, new client platforms, or times inconsistent with the user’s history. Correlate successful VPN authentication followed by unusual HTTPS requests to the VPN portal, appliance crashes, process restarts, configuration changes, account creation, certificate or trustpoint changes, changes to AAA servers, disabled logging, or unexpected management access. Because successful exploitation runs as root, appliance logs may be incomplete or manipulated; compare device configuration and software integrity against known-good backups and use external logs such as AAA servers, SIEM, NetFlow, and upstream firewalls. For threat hunting, include activity associated with Cisco ASA/FTD exploitation campaigns, ArcaneDoor-related reporting, and September 2025 active exploitation reporting, but avoid relying on a single campaign label because the vulnerability can be exploited by any actor with credentials and exploit capability.
- CVE-2025-20333
- webvpn
- AnyConnect
- Clientless SSL VPN
- Cisco Secure Firewall ASA
- Cisco Secure Firewall Threat Defense
- ASA VPN
- FTD VPN
- ArcaneDoor
- unexpected write memory
- copy running-config startup-config
- username privilege 15
- aaa-server
- no logging enable
- clear logging
- crypto ca trustpoint
What your security team should do
- Immediately inventory all Cisco ASA and FTD appliances that expose VPN web services to the internet or untrusted networks.
- Prioritize emergency upgrade of affected appliances to Cisco-fixed releases; treat KEV-listed, actively exploited internet-facing devices as highest priority.
- Require MFA for all VPN authentication and disable dormant, shared, or weakly protected VPN accounts.
- Review recent VPN authentications for unusual source countries, impossible travel, new devices, atypical usernames, or logins preceding unexpected appliance behavior.
- Restrict VPN web access by source IP or upstream controls where operationally feasible until upgrades are complete.
- Hunt for signs of post-exploitation on the appliance, including unauthorized configuration changes, new users, altered AAA settings, disabled logging, unexpected certificates, and unexplained reboots or crashes.
- Rotate credentials associated with VPN, local appliance administration, AAA integrations, and privileged accounts if exploitation is suspected.
- Preserve logs and configuration snapshots before making major changes when compromise is possible, because root-level attackers may tamper with evidence.
- Monitor Cisco, CISA KEV, and threat intelligence reporting for updated fixed versions, exploit details, and campaign-specific indicators related to Cisco ASA/FTD exploitation activity.
- If compromise is confirmed or strongly suspected, rebuild or reimage the appliance from trusted media rather than relying solely on in-place cleanup.
The record
A vulnerability in the VPN web server of Cisco Secure Firewall Adaptive Security Appliance (ASA) Software and Cisco Secure Firewall Threat Defense (FTD) Software could allow an authenticated, remote attacker to execute arbitrary code on an affected device. This vulnerability is due to improper validation of user-supplied input in HTTP(S) requests. An attacker with valid VPN user credentials could exploit this vulnerability by sending crafted HTTP requests to an affected device. A successful exploit could allow the attacker to execute arbitrary code as root, possibly resulting in the complete compromise of the affected device.
Affected packages
| Ecosystem | Package | Introduced | Fixed |
|---|