Skip to main content
Mallory
CriticalExploited in the wildCISA KEVPublic exploitVulnerability

CVE-2026-20131: Unauthenticated Root RCE in Cisco Secure Firewall Management Center

Published Mar 4, 2026 · Updated 4 days ago
Trending this weekNot trendingOutside this week’s top 100
Mention activity · 30 days9 mentions · peak 4 in a dayToday
CVSS10.0Critical · v3.1
EPSS44.4%99th percentile
CISA KEVYesAdded Mar 19, 2026
In the wildActivePer CISA KEV
Vendor
Cisco Systems
IdentifiersCVE-2026-20131CWE-502 · Deserialization of Untrusted Data

CVE-2026-20131 is an insecure Java deserialization vulnerability in the web-based management interface of Cisco Secure Firewall Management Center (FMC) Software. Deserialization of an attacker-supplied Java byte stream allows an unauthenticated remote attacker to submit a crafted serialized Java object and execute arbitrary Java code with root privileges. The vulnerability has a CVSS v3.1 score of 10.0. Interlock ransomware operators exploited it as a zero-day beginning January 26, 2026, before its public disclosure on March 4, 2026.

Do you run this?See affected versions in your environment and whether adversaries are exploiting this vulnerability against assets like yours.
Check your exposure
Analyst brief

Impact, mitigation & remediation

What it means and what to do now: patch path, interim mitigations, and the assume-compromise checklist.

Impact

What an attacker gets, and what they’ve been doing with it.

Successful exploitation permits complete compromise of the FMC appliance, affecting the confidentiality, integrity, and availability of the centralized firewall-management platform. Root access can expose stored credentials and security configurations, permit unauthorized changes, and support persistent access, data theft, and disruption. Observed Interlock intrusions used the vulnerability for initial access preceding backdoor deployment, reconnaissance, credential theft, lateral movement, and ransomware operations.

Mitigation

If you can’t patch tonight, do this now.

Remove public internet access to the FMC web-management interface and restrict connectivity to trusted administrative networks. Exposure restrictions reduce the attack surface but do not correct the deserialization flaw or protect against attackers who can already reach the interface. Review systems for suspicious activity dating back to January 26, 2026, including unauthorized remote-management software, unexpected outbound connections, and persistent implants.

Remediation

Patch, then assume compromise.

Immediately install Cisco's security updates for CVE-2026-20131 or upgrade affected FMC deployments to a fixed release specified in the applicable Cisco advisory. Cisco Security Cloud Control Firewall Management was remediated automatically and requires no customer update action. Investigate potentially exposed systems for prior compromise; patching alone does not remove attacker-established access.

Public exploits

Exploits

3 exploits validated by Mallory's analysis. Hidden: 4 rejected by Mallory's analysis as not a working exploit.

View more in app →
CVE-2026-20131-POCRepository contains a claimed working PoC exploit for CVE-2026-20131 targeting Cisco Catalyst SD-WAN Controller (vSmart) and Catalyst SD-WAN Manager (vManage). Structure: (1) `CVE-2026-20131-POC.py` is the main exploit entry point but is PyArmor-obfuscated/encrypted, preventing static extraction of exact request paths, parameters, or hardcoded targets from the provided content. (2) `cmd.jsp` is a JSP webshell payload that executes arbitrary OS commands supplied via the `cmd` HTTP parameter using `bash -c`, returning both stdout and stderr—indicative of post-exploitation persistence/command execution. (3) `README.md` describes capabilities: pre-auth auth bypass, admin privilege gain, ability to create a rogue peer in the SD-WAN control/management plane, and access to NETCONF on TCP/830. Fingerprintable observables available from the repo content are limited to the Telegram URL and the NETCONF port reference; any additional exploit endpoints (e.g., specific vManage/vSmart API paths) are likely embedded inside the obfuscated Python script and are not visible here.p3Nt3st3r-sTAr · Disclosed Mar 6, 2026 · python · jsp · network (pre-authentication remote exploitation against Cisco SD-WAN management/control plane; likely leads to admin access/RCE)OperationalCVE-2026-20131Repository contains a detection script and an RCE PoC for CVE-2026-20131 (Cisco Secure Firewall Management Center Java deserialization). Structure: (1) README.md documents the vulnerability, affected products, and usage. (2) check.py is a safe network probe that POSTs Java serialization magic bytes (0xACED0005) with Content-Type application/x-java-serialized-object to several FMC endpoints (/j_spring_security_check, /api/fmc_platform/v1/auth/generatetoken, /dispatcher, /invoker/JMXInvokerServlet, /invoker/EJBInvokerServlet) and flags HTTP 500/200 as potential deserialization handling. (3) poc.py is an operational PoC exploit that generates a malicious serialized object using an external tool (ysoserial-all.jar executed via local Java) with selectable gadget chains (CommonsCollections*, Spring*, Groovy1) and an attacker-supplied command, then POSTs the payload to likely deserialization endpoints. The exploit is unauthenticated and intended to achieve blind command execution (often inferred from HTTP 500 and/or out-of-band callbacks such as reverse shells or DNS). No persistence is implemented; payload is arbitrary command execution.sak110 · Disclosed Mar 11, 2026 · python · markdown · networkOperationalB63-POC-CVE-2026-20131The supplied tree contains 57 files, including six Python and five Java source files, Docker/Compose configuration, servlet XML, research and delivery documentation, 34 verification transcripts, an evidence index, an MIT license, and a SHA-256 manifest. No original repository URL, git reference, or archive size was supplied; empty strings and zero represent unavailable repository metadata. Several main source files and the writeup are truncated in the provided content, limiting complete implementation review. 01_research records the claimed CVE advisory and affected-version information. 02_lab implements a deliberately vulnerable Tomcat 9/JDK 17 surrogate and an ObjectInputFilter-protected twin. Its filter covers every route and deserializes request bodies identified by Java serialization magic or matching content types before normal servlet authentication. Both containers explicitly run as root. Reported versions 7.4.5 and 7.4.6 are surrogate identity labels, not independently tested Cisco releases. Gadget JARs and generated binary reference payloads are absent from the publication tree and must be supplied or regenerated. Although documentation describes pinned JAR digests, the displayed Dockerfile does not enforce those digests or automatically download the JARs. 03_exploit contains optimized and readable standalone drivers, each with a Python Java-serialization compiler. The drivers fingerprint the login surface, compare a benign serialized HashMap against equal-length zero-filled controls across nine candidate paths, compile commons-collections gadget graphs, and POST them as application/x-java-serialized-object. The optimized build adds concurrent discovery, repeat runs, and JSON output. Payload modes support arbitrary commands with returned output, bash reverse shells, and remote script staging. The compilers do not require Java or ysoserial on the operator host. Their gadget chain reaches Runtime.exec through HashSet reconstruction, TiedMapEntry hashing, LazyMap lookup, and transformer reflection. Metadata was measured against the lab JVM, not a Cisco appliance. 04_verify contains supplied execution receipts and a shell-check harness. Later raw transcripts report uid=0 command output, root-only file access, an interactive shell, stage retrieval and marker creation, 10/10 vulnerable-surrogate callbacks, and 5/5 protected-surrogate refusals. Reference comparisons show one differing HashMap threshold byte per stream, rather than literal byte identity. Earlier transcripts include failed harness runs and misleading summaries; their raw output and later corrections should take precedence. Under --no-listen, the driver cannot observe external-handler success and may report failure despite a working reverse shell. A displayed shell pipeline also masks the driver's nonzero exit status. Missing callbacks establish lack of observed proof, not necessarily absence of execution. This is genuine executable deserialization exploit material, not merely detection code or an apparent fake. However, its demonstrated scope is a purpose-built surrogate: Cisco binaries were never executed, actual appliance gadget availability and sink routes remain unknown, and the protected twin models a mitigation rather than reproducing Cisco's patch. All nine paths qualify in the lab because its filter is globally mapped; differential responses indicate candidate stream processing and do not establish exploitable gadgets or distinguish the protected twin. Advisory/version and incident-history claims are repository assertions, not externally verified findings. Fixed callback path /cve-2026-20131/beacon, Java stream magic AC ED 00 05, serialized-object content type, and User-Agents b63-fmc/1.1 and b63-fmc-probe/1.0 are useful traffic fingerprints. Callback destinations and stage URLs are operator-controlled; no fixed malicious external infrastructure is established. The Dockerfile's rm -rf operations remove container webapp/build directories and are not, by themselves, evidence of a destructive fake exploit.basesixtythree · Disclosed Oct 26, 2026 · python · java · network · webOperational
Exposure surface

Affected products & vendors

Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.

VendorProductType
Cisco SystemsFirewall Management Centerhardware
FAQ

Frequently asked questions

Is CVE-2026-20131 exploited in the wild?
Yes. As of Nov 16, 2026, Mallory has reports of exploitation in the wild, and CISA added it to KEV on Mar 19, 2026.
Is there a public exploit for CVE-2026-20131?
Yes. Mallory tracks 3 public exploits or proofs of concept for CVE-2026-20131, the earliest disclosed Mar 6, 2026.
Is there a patch for CVE-2026-20131?
Immediately install Cisco's security updates for CVE-2026-20131 or upgrade affected FMC deployments to a fixed release specified in the applicable Cisco advisory. Cisco Security Cloud Control Firewall Management was remediated automatically and requires no customer update action.
What products are affected by CVE-2026-20131?
Cisco Systems Firewall Management Center.
How severe is CVE-2026-20131?
CVE-2026-20131 has a CVSS v3.1 score of 10.0 (Critical) and an EPSS score of 44.4% (99th percentile) as of Nov 16, 2026.
Activity feed

Recent activity

239 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.

View more in app →
+22 more in the timeline. Sign up to see them
Do you run this?See affected versions in your environment and whether adversaries are exploiting this vulnerability against assets like yours.
Check your exposure
Mallory platform

The version that knows your environment.

This page is what’s public. Mallory adds the parts that aren’t: which of your assets are affected, which adversaries are exploiting it right now, and which detections to deploy tonight.
Exposure mappingQuery your assets running an affected version, and investigate the blast radius.
Threat actor evidence16Every report linking this CVE to a named adversary.
Associated malware20Malware families linked to this CVE, with evidence and IOCs.
Detection signatures1YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Vendor-by-vendor mappingCross-references every affected SKU, including bundled OEM variants.
Social activity170Community discussion across Reddit, Mastodon, and other social sources.