Skip to main content
Mallory
MalwareUsed by 2 actors

Astaroth

Also known asGuildma

Astaroth, also known as Guildma, is a Brazilian banking trojan targeting Windows systems. The content describes it as part of the Latin American banking malware ecosystem and explicitly associates it with Brazilian banker activity alongside families such as Grandoreiro, Casbaneiro, Mekotio, and Mispadu. It has been linked in reporting to TA2725, which has used Brazilian banking malware including Astaroth to target organizations mainly in Brazil, Mexico, and Spain, and to the STAC3150 cluster observed in 2025. Recent reporting in the content also describes WhatsApp-based propagation campaigns in Brazil, including a campaign dubbed Boto Cor-de-Rosa and STAC3150 activity in which WhatsApp session and contact theft was used to spread and deploy Astaroth.

Capabilities directly mentioned in the content include credential theft and banking-focused collection. Astaroth uses JavaScript for core functionality, uses WMIC to execute payloads, can be loaded through regsvr32.exe, and has used a DGA for command-and-control communications. It uses the external NirSoft tool NetPass to recover passwords, collects clipboard data via OpenClipboard and GetClipboardData, stages stolen information in a plaintext file named r1.log, Base64-encodes data before transmission, and exfiltrates collected information from r1.log to an external C2 server.

Delivery and infection vectors mentioned include malicious email attachments, phishing emails, and malicious VBS, LNK, and HTML files for execution. One detailed 2026 intrusion chain attributed with high confidence to Astaroth/Guildma used a UTF-16LE PowerShell dropper (sthzr.ps1, SHA256 4ba3c1a20ce0bdc7f78820d9bfc337cabc75087bfce2a080e15f694d29ae4715) with anti-analysis checks and forced reboot behavior. That chain downloaded staged content from meusitehostgator.com.br infrastructure, wrote artifacts including C:\Users\Public\jyyjq.txt, C:\Users\Public\cdzbc.txt, and C:\Users\Public\cxbcw_01.ps1, reflectively loaded a .NET assembly from comma-separated byte values, and abused installutil.exe from C:\Windows\Microsoft.NET\Framework\v4.0.30319\installutil to execute a final ConfuserEx-protected .NET payload hosted at https://catalogo.castrouria.com/bl.txt. The decoded final payload had SHA256 0cb0277ade2fb8bfc49c702621c31096e1268d38d7aee06436311affa06fe750 and embedded 14 hardcoded SHA256 certificate fingerprints used to identify TLS certificates for targeted Brazilian banking portals, including Itau Unibanco, Bradesco, Caixa Economica Federal, Banco do Brasil, and Santander Brasil.

The content consistently characterizes Astaroth/Guildma as a banking trojan focused on Brazilian financial targets, with campaigns extending to Brazil, Mexico, and Spain, and with both traditional phishing delivery and newer WhatsApp-based propagation mechanisms.

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.

THREAT ACTORS

Groups observed using it

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

View more details
TA2725

TA2725 is a threat actor Proofpoint tracked since March 2022 that is known for using Brazilian banking malware (including Mispadu, Astaroth, and historically Grandoreiro) and credential phishing to target organizations mainly in Brazil, Mexico, and Spain.

via proofpoint threat insight blogproofpoint.com
stac3150

...another set of attacks has led to the deployment of the Astaroth banking trojan. Sophos is tracking the second cluster under the moniker STAC3150 since September 24, 2025.

via the hacker newsthehackernews.com
MITRE ATT&CK

Techniques & procedures

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

Initial Access

2 techniques
T1566PhishingEvidence1

The attack begins with a message containing a ZIP archive ... For any user, getting a file from a friend on WhatsApp feels much safer than opening a random email.

T1566.001Spearphishing AttachmentEvidence2

The attack begins with a message containing a ZIP archive (basically a compressed folder), usually named with a confusing string of digits like 552_516107-a9af16a8-552.zip.

Execution

8 techniques
T1047Windows Management InstrumentationEvidence1
TacticExecution

The content repeatedly describes threat actors and malware using WMI/WMIC/wmiexec for remote execution, lateral movement, discovery, persistence, and administrative actions; e.g., 'APT41 used WMI in several ways, including for execution of commands via WMIEXEC as well as for persistence via PowerSploit' and 'Scattered Spider used Windows Management Instrumentation (WMI) to move laterally via Impacket.'

T1059.001PowerShellEvidence1
TacticExecution

The dropper is a UTF-16LE PowerShell script... The sample sthzr.ps1 arrived as a 6KB UTF-16LE encoded PowerShell file.

T1059.003Windows Command ShellEvidence2
TacticExecution

During the 2016 Ukraine Electric Power Attack, Sandworm Team used the xp_cmdshell command in MS-SQL. During the 2025 Poland Wiper Attacks, the adversaries leveraged PsExec to run cmd.exe commands on multiple victim machines. Numerous malware families and groups are described as using cmd.exe, cmd /c, Windows command shell, or command-line interfaces to execute commands, payloads, reconnaissance, persistence, cleanup, and ransomware actions.

T1059.005Visual BasicEvidence2
TacticExecution

If a victim opens this folder, a hidden script triggers a chain reaction.

T1059.006PythonEvidence1
TacticExecution

The WhatsApp Spreader: This is a new piece of code written in Python (a file named zapbiu.py) that steals your contact list and starts sending out copies of the virus to everyone you know.

T1059.007JavaScriptEvidence1
TacticExecution

AppleSeed has the ability to use JavaScript to execute PowerShell. APT32 has used JavaScript for drive-by downloads and C2 communications. Astaroth uses JavaScript to perform its core functionalities.

T1204User ExecutionEvidence2
TacticExecution

If a victim opens this folder, a hidden script triggers a chain reaction.

T1204.002Malicious FileEvidence1
TacticExecution

Examples include: "Sandworm Team leveraged Microsoft Office attachments which contained malicious macros..."; "Bumblebee has relied upon a user opening an ISO file to enable execution of malicious shortcut files and DLLs"; "Lumma Stealer has gained initial execution through victims opening malicious executable files embedded in zip archives, and MSI files within RAR files."

Persistence

2 techniques
T1547.001Registry Run Keys / Startup FolderEvidence3

The content repeatedly describes malware and threat actors establishing persistence by adding values under HKCU/HKLM\Software\Microsoft\Windows\CurrentVersion\Run or RunOnce, and by placing executables, scripts, .lnk files, or .bat files in the Windows Startup folder.

T1547.009Shortcut ModificationEvidence1

The content repeatedly references malicious shortcut files: e.g., "APT38 has used malicious Word documents and shortcut files," "Bumblebee... opening an ISO file to enable execution of malicious shortcut files and DLLs," and "Mustang Panda distributed malicious LNK objects for user execution."

T1055Process InjectionEvidence1

The assembly imports... Win32 injection APIs (LoadLibrary, GetProcAddress, GetDelegateForFunctionPointer, OpenProcess) -- the toolkit for process injection

T1547.001Registry Run Keys / Startup FolderEvidence3

The content repeatedly describes malware and threat actors establishing persistence by adding values under HKCU/HKLM\Software\Microsoft\Windows\CurrentVersion\Run or RunOnce, and by placing executables, scripts, .lnk files, or .bat files in the Windows Startup folder.

T1547.009Shortcut ModificationEvidence1

The content repeatedly references malicious shortcut files: e.g., "APT38 has used malicious Word documents and shortcut files," "Bumblebee... opening an ISO file to enable execution of malicious shortcut files and DLLs," and "Mustang Panda distributed malicious LNK objects for user execution."

Stealth

11 techniques
T1027Obfuscated Files or InformationEvidence4
TacticStealth

The content repeatedly describes payloads, strings, configuration files, scripts, URLs, and binaries being obfuscated or encoded using Base64, XOR, RC4, AES, RSA, hex encoding, custom algorithms, and other methods across many malware families and threat actors.

T1027.003SteganographyEvidence1
TacticStealth

Later, in October 2025, it was found abusing GitHub to hide its backup files inside images.

T1027.010Command ObfuscationEvidence1
TacticStealth

That string concatenation is the second layer of evasion within the evasion... jTifh -- a function that executes: "Resta" + "rt-Computer -Force".

T1027.013Encrypted/Encoded FileEvidence1
TacticStealth

Examples throughout the content include 'encrypted payloads decrypted and executed in memory,' 'encrypts its configuration file,' 'AES-encrypted resource,' 'RC4 encrypted embedded scripts,' and 'payload includes an encrypted main component.'

T1055Process InjectionEvidence1

The assembly imports... Win32 injection APIs (LoadLibrary, GetProcAddress, GetDelegateForFunctionPointer, OpenProcess) -- the toolkit for process injection

T1140Deobfuscate/Decode Files or InformationEvidence1
TacticStealth

The content repeatedly describes malware and threat actors decoding, decrypting, or deobfuscating payloads, strings, configuration data, commands, and C2 traffic prior to execution or use, e.g., 'APT28 macro uses the command certutil -decode to decode contents of a .txt file storing the base64 encoded payload' and 'Action RAT can use Base64 to decode actor-controlled C2 server communications.'

T1218.004InstallUtilEvidence1
TacticStealth

installutil.exe is a Microsoft-signed .NET Framework utility... Because it is signed by Microsoft, it is trusted by default in AppLocker and WDAC policies... The malware abuses this trust to load arbitrary .NET code.

T1497.001System ChecksEvidence1

Before executing any payload logic, the dropper verifies internet connectivity... In a sandbox that intercepts or blocks ICMP, the result is either $null or a single error object... If connectivity passes, the dropper checks for running analysis tools.

T1497.003Time Based ChecksEvidence1

MITRE ATT&CK ID Technique Implementation... T1497.003 Time Based Evasion Start-Sleep -Seconds 5 delays between stages

T1564.001Hidden Files and DirectoriesEvidence1
TacticStealth

Further probing revealed that the malware hides its main files in a very specific spot on the computer: C:\Public\MicrosoftEdgeCache_6.60.2.9313.

T1620Reflective Code LoadingEvidence1
TacticStealth

Calls [System.Reflection.Assembly]::Load($bytes) -- reflective loading that avoids writing the DLL to disk as a PE file

T1553Subvert Trust ControlsEvidence1

The decoded DLL embeds 14 SHA256 certificate fingerprints hardcoded as string constants... When Astaroth detects a TLS handshake whose server certificate matches one of these hashes, it interposes on the connection to capture banking credentials.

Credential Access

5 techniques
T1528Steal Application Access TokenEvidence1

In February 2025, a version of Astaroth was found that could bypass two-factor authentication to steal Gmail and Microsoft logins.

T1539Steal Web Session CookieEvidence1

MITRE ATT&CK ID Technique Implementation... T1539 Steal Web Session Cookie Browser injection for banking session credential capture

T1552Unsecured CredentialsEvidence1

MITRE ATT&CK ID Technique Implementation... T1552 Unsecured Credentials Banking credential theft via overlay/hook on targeted HTTPS sessions

T1555Credentials from Password StoresEvidence2

Agent Tesla has the ability to steal credentials from FTP clients and wireless profiles... APT33 has used a variety of publicly available tools like LaZagne to gather credentials... Mimikatz performs credential dumping to obtain account and password information useful in gaining access to additional systems and enterprise network resources. It contains functionality to acquire information about credentials in many ways, including from the credential vault and DPAPI.

T1649Steal or Forge Authentication CertificatesEvidence1

In February 2025, a version of Astaroth was found that could bypass two-factor authentication to steal Gmail and Microsoft logins.

Discovery

6 techniques
T1016System Network Configuration DiscoveryEvidence1
TacticDiscovery

The content repeatedly describes actors and malware using commands and APIs such as ipconfig /all, ifconfig, arp -a, route print, netsh interface show, GetAdaptersInfo, and GetIpNetTable to gather IP addresses, MAC addresses, DNS, DHCP, gateways, routing tables, ARP cache, proxy settings, and network adapter/interface details.

T1057Process DiscoveryEvidence1
TacticDiscovery

The content repeatedly describes malware and threat actors obtaining lists of running processes, using utilities such as tasklist, ps, WMI, Get-Process, CreateToolhelp32Snapshot, EnumProcesses, and similar APIs/commands to enumerate active processes on victim systems.

T1082System Information DiscoveryEvidence1
TacticDiscovery

The content repeatedly describes malware and threat actors collecting host details such as OS version, hostname, architecture, CPU, memory, BIOS, domain, language, and other configuration data; e.g., "APT41 uses multiple built-in commands such as systeminfo and net config Workstation to enumerate victim system basic configuration information."

T1497.001System ChecksEvidence1

Before executing any payload logic, the dropper verifies internet connectivity... In a sandbox that intercepts or blocks ICMP, the result is either $null or a single error object... If connectivity passes, the dropper checks for running analysis tools.

T1497.003Time Based ChecksEvidence1

MITRE ATT&CK ID Technique Implementation... T1497.003 Time Based Evasion Start-Sleep -Seconds 5 delays between stages

T1614.001System Language DiscoveryEvidence1
TacticDiscovery

Examples include "Bazar can also check if the Russian language is installed on the infected machine and terminate if it is found," "DropBook has checked for the presence of Arabic language," and "Maze has checked the language of the infected system using the GetUSerDefaultUILanguage function."

Collection

1 technique
T1074Data StagedEvidence1

The content repeatedly describes adversaries and malware storing collected data, command output, credentials, archives, or files in local temporary folders, working directories, hidden directories, registry locations, recycle bins, or specific files prior to exfiltration.

T1071.001Web ProtocolsEvidence1

The dropper configures its web client... Headers.Add("Cache-Control", "no-cache")... The secondary URL decodes to https://catalogo.castrouria.com/bl.txt

T1132Data EncodingEvidence2

C2 traffic from ADVSTORESHELL is encrypted, then encoded with Base64 encoding... APT19 HTTP malware variant used Base64 to encode communications to the C2 server... APT33 has used base64 to encode command and control traffic.

Exfiltration

1 technique
T1041Exfiltration Over C2 ChannelEvidence2

ADVSTORESHELL exfiltrates data over the same channel used for C2... Agrius exfiltrated staged data using tools such as Putty and WinSCP, communicating with command and control servers... numerous malware and groups sent victim data, files, credentials, or host information over existing C2 channels.

Impact

1 technique
T1529System Shutdown/RebootEvidence1
TacticImpact

If the check fails, the dropper calls jTifh... 'Restart-Computer -Force'... If any of these twelve processes are found, the same forced reboot is triggered.

INDICATORS OF COMPROMISE

IOCs tracked for this family

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

View more in app
Network
17 tracked

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

Hashes
40 tracked

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

Other
1 tracked

Other indicator types observed in public reporting.

TypeValueLatest sighting
domain●●●●●●●●●●●●View more in app3 months ago
hash.md5●●●●●●●●●●●●View more in app3 months ago
hash.sha256●●●●●●●●●●●●View more in app3 months ago
hash.sha256●●●●●●●●●●●●View more in app3 months ago
hash.sha256●●●●●●●●●●●●View more in app3 months ago
hash.sha256●●●●●●●●●●●●View more in app3 months ago
ACTIVITY FEED

Recent activity

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

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 matching58

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

Threat actor attribution2

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

Exploited vulnerabilities

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 mapping38

Every documented technique, ranked by evidence weight.

Researcher chatter

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