Skip to main content
Mallory
CriticalPublic exploitVulnerability

CVE-2024-3094: XZ Utils Backdoor

Published Mar 29, 2024 · Updated 3 days ago
Trending this weekNot trendingOutside this week’s top 100
Mention activity · 30 days7 mentions · peak 2 in a dayToday
CVSS10.0Critical · v3.1
EPSS90.9%100th percentile
CISA KEVNoNot listed
In the wildReported3 exploitation records in Mallory
Vendor
Tukaani
IdentifiersCVE-2024-3094CWE-506 · Embedded Malicious Code
Also known asliblzma_backdoorxz_backdoor

CVE-2024-3094 is a software supply-chain backdoor in XZ Utils release tarballs 5.6.0 and 5.6.1. Obfuscated build logic reconstructs a precompiled malicious object from disguised test data and incorporates it into liblzma under selected build conditions. The implant abuses glibc IFUNC resolution to redirect cryptographic functions, principally RSA_public_decrypt, when the compromised library is loaded into sshd, including indirectly through libsystemd. A specially crafted SSH authentication exchange can activate attacker-controlled code before normal authentication completes. The backdoor supports remote command execution and authentication subversion on susceptible Linux installations; simply having an affected XZ version does not establish that the SSH exploitation path is active.

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 can enable unauthenticated remote command execution with administrator privileges, potentially compromising the confidentiality, integrity, and availability of the affected host. The implant can manipulate SSH authentication processing and suppress or alter selected authentication logs, complicating detection and investigation. Compromised liblzma builds also undermine the integrity of applications that load the library, although the documented runtime activation targets sshd.

Mitigation

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

Until trusted replacement packages are installed, stop affected SSH services or isolate affected systems. If SSH must remain available, restrict network access to trusted administrative sources as a temporary exposure-reduction measure. Inventory installed XZ Utils and liblzma packages, including those in container images, and distinguish package presence from an actively backdoored build. Hunt for suspicious SSH behavior using package inspection, file and memory scanning, and behavioral detections. Do not rely exclusively on authentication logs because the implant can suppress or alter them.

Remediation

Patch, then assume compromise.

Replace XZ Utils and liblzma 5.6.0 or 5.6.1 with trusted, distribution-provided uncompromised packages. Identified rollback options include upstream 5.4.6 or the appropriate vendor-maintained 5.4.x package. Debian identified 5.6.1+really5.4.5-1 as its reverted fixed package. Follow distribution-specific recovery guidance and restart affected processes or reboot to eliminate loaded compromised libraries. Investigate systems that ran affected builds for unauthorized access and persistence; if compromise is established or system integrity cannot be verified, rebuild from trusted media and rotate exposed credentials and keys.

Public exploits

Exploits

5 exploits validated by Mallory's analysis. Hidden: 55 rejected by Mallory's analysis as not a working exploit; 3 not yet analyzed, such as Exploit-DB entries.

View more in app →
ludus_xz_backdoorThis repository is an Ansible role designed to automate the installation (and optional removal) of the xz backdoor (CVE-2024-3094) on Debian-based Linux systems. The role downloads a malicious version of the liblzma shared library from a remote GitHub repository and replaces the system's liblzma.so.5 with a symlink to the malicious version, thereby activating the backdoor. It can also install the xzbot command-line tool, which is used to interact with the backdoor. The role supports both installation and uninstallation (restoring the original library). The main logic is implemented in Ansible YAML files under the 'tasks' directory, with 'tasks/main.yml' serving as the entry point. The attack vector is local, as the role must be executed with administrative privileges on the target system. The repository is intended for use in controlled environments (e.g., Ludus cyber ranges) for demonstration or research purposes, but it deploys real malware and should be handled with caution. Notable endpoints include the URLs for downloading the malicious library and the file paths where the backdoor and tools are installed.badsectorlabs · Disclosed Apr 5, 2024 · yaml · ansible · localOperationalCVE-2024-3094This repository contains a single C-language exploit (exploit.c) targeting the OpenSSH 9.2p1 server vulnerability known as 'regreSSHion' (CVE-2024-6387). The exploit leverages a race condition in the SSH daemon's signal handler, allowing remote code execution as root on glibc-based Linux systems. The exploit is operational and requires the attacker to specify the target's IP address and SSH port. The code includes a shellcode placeholder, which the user is expected to replace with a custom payload. The exploit attempts to brute-force the race condition, iterating over possible glibc base addresses and performing thousands of attempts as recommended in the advisory. The README provides basic usage instructions. No hardcoded IPs, domains, or file paths are present; the exploit is generic and requires user-supplied target information. The repository is focused and contains only the exploit code, a README, and a .gitattributes file.DANO-AMP · Disclosed Jul 5, 2024 · c · networkOperationalcve-2024-3094This repository provides a comprehensive toolkit for analyzing, exploiting, and detecting the XZ Utils (liblzma) backdoor (CVE-2024-3094) affecting versions 5.6.0 and 5.6.1. The structure includes: - **bot.go**: A Go-based exploit tool that crafts a malicious SSH authentication request, leveraging the backdoored liblzma to execute arbitrary commands on the target system. It connects to a specified SSH address (default 127.0.0.2:2222) and sends a payload that triggers the backdoor. - **reverse_shell.py**: A Python script that can interact with the backdoor, acting as a custom SSH agent to facilitate further exploitation, such as establishing a reverse shell. - **patch.py** and **patch_libzma.py**: Scripts to patch the vulnerable liblzma shared object, replacing the backdoor's key generation routine with a static key, making exploitation deterministic for testing or demonstration. - **generator.py**: Generates Ed448 key pairs used in the exploit process. - **detector.sh**: A Bash script to detect if the system's sshd is linked to a vulnerable (backdoored) liblzma library by checking for specific function signatures. - **configure.sh** and **inject.sh**: Scripts to automate the setup of a vulnerable environment and inject the malicious library into the system. - **rule.yar**: YARA rule for detecting the malicious code in liblzma binaries. The repository is intended for educational and defensive research. It demonstrates how the backdoor can be exploited to bypass SSH authentication and execute arbitrary commands, and provides tools for both exploitation and detection. The main attack vector is network-based, targeting SSH services linked against the compromised liblzma library. The exploit requires the attacker to know the specific cryptographic keys used by the backdoor, which are generated or patched using the provided scripts.valeriot30 · Disclosed Jun 3, 2025 · go · python · networkOperationalxzbotThis repository provides a comprehensive toolkit for exploring and exploiting the xz/liblzma backdoor (CVE-2024-3094). The main components are: - `main.go`: Implements a Go-based CLI tool (`xzbot`) that crafts and sends a specially formatted SSH certificate to a vulnerable SSH server, triggering the backdoor and executing an arbitrary shell command (default: `id > /tmp/.xz`). The tool allows configuration of the target address, the ED448 key seed, and the command to execute. - `patch.py`: A Python script that patches a backdoored `liblzma.so.5.6.1` shared object, replacing the embedded ED448 public key with one generated from a known seed, allowing the user to exploit the backdoor with their own key. - `openssh.patch`: A patch for OpenSSH that adds logging for connections attempting to exploit the backdoor, useful for honeypot or detection purposes. - `README.md`: Provides detailed technical documentation, including the backdoor's cryptographic format, usage instructions, and references. The exploit works by sending a malicious SSH certificate with a payload in the CA signing key field, which the backdoored liblzma interprets as a command to execute via system(). The repository also includes detection and research tools, such as a honeypot patch for OpenSSH. The exploit is operational and can be used to achieve remote code execution on vulnerable systems, provided the attacker controls the ED448 key used by the backdoor.amlweems · Disclosed Apr 1, 2024 · go · python · networkPoCxzutils_backdoor_obfuscationThis repository provides a Bash script ('obfuscate_payload.sh') that recreates the obfuscation technique used in the XZ Utils supply chain attack (CVE-2024-3094). The script takes a user-supplied payload, compresses and obfuscates it, and embeds it into a modified 'good-large_compressed.lzma' file. This file is intended to be placed in the 'test/files' directory of the xzutils 5.6.0 source code. When xzutils is built, the backdoor payload is extracted and executed, enabling arbitrary code execution on the build system. The repository includes a README with detailed instructions for both obfuscation and deobfuscation, as well as a LICENSE file. The exploit is a proof-of-concept for demonstrating the obfuscation and delivery technique used in the real-world XZ backdoor incident. No network endpoints are present; all operations are local file manipulations and build process hijacking.ThomRgn · Disclosed Oct 22, 2025 · bash · localPoC
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
TukaaniXzapplication
FAQ

Frequently asked questions

Is CVE-2024-3094 exploited in the wild?
Yes. As of Nov 19, 2026, Mallory has reports of exploitation in the wild.
Is there a public exploit for CVE-2024-3094?
Yes. Mallory tracks 5 public exploits or proofs of concept for CVE-2024-3094, the earliest disclosed Apr 1, 2024.
Is there a patch for CVE-2024-3094?
Replace XZ Utils and liblzma 5.6.0 or 5.6.1 with trusted, distribution-provided uncompromised packages. Identified rollback options include upstream 5.4.6 or the appropriate vendor-maintained 5.4.x package.
What products are affected by CVE-2024-3094?
Tukaani Xz.
How severe is CVE-2024-3094?
CVE-2024-3094 has a CVSS v3.1 score of 10.0 (Critical) and an EPSS score of 90.9% (100th percentile) as of Nov 19, 2026.
Activity feed

Recent activity

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

View more in app →
medium the sleuth sheetNews · Nov 16, 2026Espionage & AI: ARTEX, Building Agentic Hacking Tools | by VEEXH | The Sleuth Sheet | Nov, 2026 | MediumMentioned only in an illustrative model prompt. The content provides no technical description, affected products, or exploitation details for this vulnerability.Read more ↗codebyNews · Nov 9, 2026Заброшенный домен → RCE: supply chain атака за $10A malicious backdoor introduced into XZ Utils through a prolonged supply-chain compromise. The article states that modified liblzma reached testing versions of Linux distributions and cites a CVSS score of 10.0 and classification CWE-506. It appears as a background comparison, not as the article's primary subject.Read more ↗medium the sleuth sheetNews · Oct 30, 2026Espioange & AI: Mastering the Criminal Mindset | by VEEXH | The Sleuth Sheet | Oct, 2026 | MediumUnknown. The vulnerability appears only in an illustrative AI prompt; the content provides no technical details or evidence of exploitation.Read more ↗malware newsNews · Oct 19, 2026What Is a Backdoor? Meaning, Types, Real Examples, Detection & Removal (2026 Guide) - Malware News - Malware Analysis, News and IndicatorsA supply-chain backdoor in XZ Utils versions 5.6.0 and 5.6.1 designed to bypass SSH authentication on affected Linux systems. The article assigns it a CVSS score of 10.0 and explains that it was discovered before the compromised versions reached stable enterprise distributions.Read more ↗
+33 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 evidence1Every report linking this CVE to a named adversary.
Associated malware6Malware families linked to this CVE, with evidence and IOCs.
Detection signaturesYARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Vendor-by-vendor mappingCross-references every affected SKU, including bundled OEM variants.
Social activity72Community discussion across Reddit, Mastodon, and other social sources.