Skip to main content
Meet us at Black Hat USA 2026— Las Vegas, August 1–6Book a Meeting
Mallory
MalwareUsed by 8 actorsExploits 2 CVEs

SparkRAT

SparkRAT is an open-source, Go-based remote access trojan/backdoor first seen in 2023 and developed by the Chinese-speaking developer XZB-1248. It is cross-platform, with variants or support for Windows, Linux, and macOS, and communicates with command-and-control infrastructure over WebSocket. Reported capabilities include direct command execution and command-line interaction, file upload and download, system fingerprinting, file and process manipulation, system manipulation, and information theft; some reporting also describes it as an infostealer. The malware includes an automatic upgrade mechanism that sends an HTTP POST request containing a commit query parameter representing the current version. Multiple reports note per-target SparkRAT builds distinguished by differing ldflags COMMIT values.

SparkRAT has been consistently associated with several intrusion clusters and campaigns. SentinelLABS described the DragonSpark activity in East Asia as the first concrete malicious campaign showing consistent use of SparkRAT and assessed the operator as a Chinese-speaking threat actor. SparkRAT was also reported in activity linked to Webworm, RedNovember, TAG-140, and TGR-STA-1030/UNC6619, and has been referenced in various Chinese-linked campaigns. In TGR-STA-1030 reporting, SparkRAT appeared alongside Cobalt Strike, VShell, Havoc, and Sliver; in TAG-140 reporting it was one of several RAT payloads used against Indian targets.

Observed delivery and execution methods vary by campaign. Kroll documented a Golang loader named LESLIELOADER used to decode, decrypt, and inject SparkRAT into a suspended notepad.exe process; the loader used Base64 decoding and AES-192 decryption with the key string "LeslieCheungKwok," and in one case attempted to beacon to 209.141.50[.]215:443. Other reporting states LESLIELOADER downloaded SparkRAT, and Huntress observed SparkRAT deployed after exploitation of IIS web application flaws. SparkRAT has also been deployed in exploitation of public-facing vulnerabilities, including Apache ActiveMQ CVE-2023-46604 and BeyondTrust Remote Support / Privileged Remote Access CVE-2026-1731.

In the BeyondTrust CVE-2026-1731 exploitation wave, multiple sources reported SparkRAT and VShell being deployed after unauthenticated OS command injection against internet-exposed remote access appliances. Affected sectors included financial services, technology, higher education, legal services, healthcare, retail, and other organizations across the United States, France, Germany, Australia, and Canada. Post-exploitation activity in those cases included reconnaissance, web shell deployment, persistence, lateral movement, remote management tool installation, tunneling, and in some cases data theft.

High-confidence indicators and technical details directly mentioned in the content include SparkRAT’s WebSocket C2, cross-platform support, open-source Go implementation, DragonSpark-linked usage, per-victim COMMIT-tagged builds, and one observed DragonSpark sample version built from commit 6920f726d74efb7836a03d3acfc0f23af196765e on 2022-11-01 UTC.

Share:
For your environment

Hunt this family in your stack

Mallory pivots from this family to the IOCs, detections, and named campaigns that touch your stack, and pages you when something new lands.

EXPLOITED CVES

Vulnerabilities exploited

2 CVEs Mallory has correlated with this family across public research and vendor advisories. Each row links to the full Mallory page for that vulnerability.

2 CVES
CVE-2026-1731Pre-auth OS Command Injection RCE in BeyondTrust Remote Support and Privileged Remote AccessExploited in the wild

The vulnerability, tracked as CVE-2026-1731, is an operating system command injection flaw that also impacts some older versions of BeyondTrust Privileged Remote Access. The flaw allows an attacker to execute arbitrary commands on a server without the need for credentials or any user interaction. | A critical vulnerability in BeyondTrust Remote Support is facing an increase in threat activity, with hackers deploying SparkRAT and vShell backdoors and using remote management tools to conduct reconnaissance...

via cybersecurity divecybersecuritydive.com
CVE-2024-43451Microsoft Windows NTLM Hash Disclosure Spoofing VulnerabilityExploited in the wild

ClearSky Cyber Security has uncovered a new zero-day vulnerability, CVE-2024-43451, actively exploited in the wild, targeting Windows systems primarily in Ukraine. This flaw enables attackers to exploit URL files for malicious activity by performing actions as simple as a single right-click.

via security online infosecurityonline.info
THREAT ACTORS

Groups observed using it

8 distinct threat actors attributed by public researchers. Open in Mallory to see the full evidence chain and overlapping campaigns.

View more details
Webworm

Sandbox family search for family:sparkrat . Returned a per-victim SparkRAT sibling submitted independently to the sandbox. Its ldflags COMMIT differs from the case sample. Confirms the operator uses per-target SparkRAT builds.

via github gist webgist.github.com
UAC-0194

The initial analysis showed that the ZIP files downloaded were installing SparkRAT on some systems, while later variations utilized Redline Stealer.

via security online infosecurityonline.info
RedNovember

...Leslieloader that downloads a backdoor dubbed SparkRAT. The Go variants are compliant with Windows, Linux and OSX. They support file upload and download, system fingerprinting and direct command-line interaction with infected hosts.

via bank info securitybankinfosecurity.com
SideCopy

"...including CurlBack, SparkRAT, AresRAT, Xeno RAT, AllaKore, and ReverseRAT."

via dark readingdarkreading.com
DragonSpark

The attacks are characterized by the use of the little known open source SparkRAT and malware that attempts to evade detection through Golang source code interpretation.

via sentinelone labssentinelone.com
unk_coltcentury

...UNK_ColtCentury... likely an attempt to deploy the SparkRAT backdoor.

via proofpoint threat insight blogproofpoint.com
MITRE ATT&CK

Techniques & procedures

18 distinct techniques documented for this family, organized by ATT&CK tactic.

Initial Access

1 technique
T1190Exploit Public-Facing ApplicationEvidence6

A critical vulnerability in BeyondTrust Remote Support is facing an increase in threat activity ... Multiple BeyondTrust Remote Support users have been confirmed targets ... The flaw allows an attacker to execute arbitrary commands on a server without the need for credentials or any user interaction.

Execution

4 techniques
T1059Command and Scripting InterpreterEvidence1

This version supports 26 commands... including execution of arbitrary Windows system and PowerShell commands.

T1059.001PowerShellEvidence1

This version supports 26 commands that implement a wide range of functionalities: Command execution: including execution of arbitrary Windows system and PowerShell commands.

T1059.004Unix ShellEvidence1

The vulnerability, tracked as CVE-2026-1731, is an operating system command injection flaw ... The flaw allows an attacker to execute arbitrary commands on a server without the need for credentials or any user interaction.

T1203Exploitation for Client ExecutionEvidence1

"CVE-2026-1731 is an OS command injection vulnerability (CWE-78) in the thin-scc-wrapper component, which is exposed directly to the network via WebSocket... lets attackers run system commands with no login required."

Persistence

1 technique
T1547Boot or Logon Autostart ExecutionEvidence1

“Unit 42 confirmed the flaw is being actively exploited for… backdoor installation…”

Privilege Escalation

1 technique
T1547Boot or Logon Autostart ExecutionEvidence1

“Unit 42 confirmed the flaw is being actively exploited for… backdoor installation…”

Stealth

1 technique
T1070.004File DeletionEvidence1

File and process manipulation: including process termination as well as file upload, download, and deletion.

Discovery

3 techniques
T1057Process DiscoveryEvidence1

Information theft: including exfiltration of platform information... and process and file enumeration.

T1082System Information DiscoveryEvidence1

Information theft: including exfiltration of platform information (CPU, network, memory, disk, and system uptime information)

T1083File and Directory DiscoveryEvidence1

Information theft: including exfiltration of platform information... and process and file enumeration.

Lateral Movement

2 techniques
T1021Remote ServicesEvidence3

Key initial access vectors include ... Exposed Outlook Web Access (OWA) and VPN infrastructure

T1210Exploitation of Remote ServicesEvidence1

“By sending specially crafted requests, an unauthenticated remote attacker may be able to execute operating system commands in the context of the site user.”

Collection

1 technique
T1113Screen CaptureEvidence1

Information theft: including exfiltration of platform information... screenshot theft

Command and Control

4 techniques
T1071Application Layer ProtocolEvidence1

"Among the tools put to use by the threat actor are command-and-control (C2) frameworks... Cobalt Strike, VShell, Havoc, Sliver, and SparkRAT"

T1071.001Web ProtocolsEvidence1

SparkRAT uses the WebSocket protocol to communicate with the C2 server and features an upgrade system.

T1105Ingress Tool TransferEvidence3

After gaining access to environments, the threat actor conducted a variety of malicious activities, such as lateral movement, privilege escalation, and deployment of malware and tools hosted at attacker-controlled infrastructure.

T1219Remote Access ToolsEvidence2

Sandbox family search for family:sparkrat . Returned a per-victim SparkRAT sibling submitted independently to the sandbox. Its ldflags COMMIT differs from the case sample. Confirms the operator uses per-target SparkRAT builds.

Impact

1 technique
T1489Service StopEvidence1

File and process manipulation: including process termination as well as file upload, download, and deletion.

INDICATORS OF COMPROMISE

IOCs tracked for this family

11 indicators attributed across vendor reports, sandbox runs, and researcher write-ups. Full values are available in Mallory.

View more in app
Network
5 tracked

IPs, domains, and DNS infrastructure linked to this family.

Hashes
5 tracked

File hashes (MD5, SHA-1, SHA-256) from samples and reports.

Other
1 tracked

Other indicator types observed in public reporting.

TypeValueLatest sighting
hash.sha1●●●●●●●●●●●●View more in app1 month ago
hash.sha256●●●●●●●●●●●●View more in app1 month ago
hash.sha256●●●●●●●●●●●●View more in app1 month ago
hash.sha1●●●●●●●●●●●●View more in app1 year ago
ip.v4●●●●●●●●●●●●View more in app1 year ago
ip.v4●●●●●●●●●●●●View more in app1 year ago
What this page doesn’t show

The version that knows your environment.

This page is what’s public. Mallory adds the parts that aren’t: which of your assets match these IOCs, which detections are missing, which campaigns to expect next, and what to do in the next 30 minutes.
IOC matching11

Match every observed IP, domain, and hash against your live telemetry.

Threat actor attribution8

Named campaigns wielding this family, with evidence pinned to each claim.

Exploited vulnerabilities2

CVEs this family uses for access and lateral movement.

Detection signatures

YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.

MITRE ATT&CK mapping18

Every documented technique, ranked by evidence weight.

Researcher chatter

Reddit, Mastodon, and CTI community discussion around this family.