ThreatBeaconXSubscribe
CriticalRansomware · 9 min read

From Bing Search to Ransomware: Bumblebee and AdaptixC2 Deliver Akira

Attackers used SEO poisoning to distribute a trojanized ManageEngine OpManager installer, delivering Bumblebee malware and later deploying AdaptixC2 for command and control. The intrusion involved privileged account creation, credential theft, RustDesk persistence, and data exfiltration through SFTP before Akira ransomware was deployed. The incident highlights the risks of malicious software downloads and the importance of detecting attacker activity before ransomware execution.

Written by Battula Bhanu Prakash·Published Oct 26, 2026

Overview

A July 2025 intrusion investigated by The DFIR Report showed how a malicious software download could become the entry point for a ransomware attack.

The incident began when a user searched Bing for “ManageEngine OpManager” and visited a malicious website, opmanager[.]pro. The website distributed a trojanized ManageEngine-OpManager.msi installer. A malicious DLL named msimg32.dll was subsequently loaded through consent.exe, resulting in the execution of Bumblebee malware.

The attackers later deployed AdaptixC2 to maintain command-and-control access and perform internal reconnaissance. Their activity included creating privileged accounts, accessing a domain controller, extracting credentials, installing RustDesk for remote access, and attempting to obtain credentials from Veeam infrastructure.

The intrusion ultimately resulted in data exfiltration and the deployment of Akira ransomware. Investigators also observed a subsequent intrusion involving RustDesk and activity against a child domain controller.

The incident highlights the risks associated with downloading software from search-engine results and the importance of detecting attacker activity before ransomware execution.

Attack Timeline

July 2025 — Initial access: A user searched Bing for “ManageEngine OpManager” and visited opmanager[.]pro, which distributed the trojanized ManageEngine-OpManager.msi installer.

Initial execution — Bumblebee deployment: The malicious DLL msimg32.dll was loaded through consent.exe, allowing Bumblebee to execute.

Approximately five hours later — AdaptixC2: The attackers deployed an AdaptixC2 beacon named AdgNsy.exe, communicating with 172[.]96[.]137[.]160:443.

Internal reconnaissance: The attackers used commands including systeminfo, nltest /dclist:, whoami /groups, and net group domain admins /dom to investigate the environment and identify privileged accounts and domain infrastructure.

Privilege escalation: The attackers created backup_DA and backup_EA accounts and added backup_EA to the Enterprise Administrators group.

Credential access and backup discovery: The attackers accessed a domain controller, used wbadmin.exe to back up NTDS.dit and the SYSTEM and SECURITY registry hives, and attempted to extract credentials from LSASS.

Persistence and lateral movement: RustDesk was installed on compromised systems, and an SSH reverse tunnel was established to 193[.]242[.]184[.]150.

Data exfiltration: FileZilla was used to transfer data through SFTP to 185[.]174[.]100[.]203.

Ransomware deployment: The attackers deployed the Akira ransomware payload locker.exe.

Subsequent activity: Two days later, the attackers returned through RustDesk, accessed a child domain controller, performed further reconnaissance, and deployed Akira in the child domain.

The reported timeline shows how the intrusion progressed from a malicious software download to domain-level compromise and ransomware deployment.

Initial Access: SEO Poisoning and Bumblebee

The initial access technique relied on search-engine manipulation rather than a direct exploitation of a software vulnerability.

The victim searched for “ManageEngine OpManager” on Bing and visited opmanager[.]pro, a malicious website that offered a trojanized installer named ManageEngine-OpManager.msi.

The installer was associated with the following SHA-256 hash:

186b26df63df3b7334043b47659cba4185c948629d857d47452cc1936f0aa5da

During execution, a malicious DLL named msimg32.dll was loaded through consent.exe. The DLL associated with the DFIR investigation had this SHA-256 hash:

a6df0b49a5ef9ffd6513bfe061fb60f6d2941a440038e2de8a7aeb1914945331

This activity led to Bumblebee malware execution and provided the attackers with a foothold in the victim environment.

The incident illustrates why defenders should not rely exclusively on website reputation or the apparent legitimacy of software filenames. Malicious installers can impersonate legitimate applications and use DLL-loading techniques to execute additional payloads.

Detection opportunities:

  • Investigate unexpected software installers downloaded from unfamiliar domains.

  • Monitor DLL loads involving unusual or untrusted application paths.

  • Correlate installer execution with newly created processes and outbound network connections.

  • Investigate endpoints communicating with known incident-related infrastructure.

AdaptixC2 Deployment and Internal Reconnaissance

Approximately five hours after the initial Bumblebee execution, the attackers deployed an AdaptixC2 beacon named AdgNsy.exe.

The observed command-and-control endpoint was:

172[.]96[.]137[.]160:443

The attackers used built-in Windows commands to collect information about the compromised environment.

Reported commands included:

  • systeminfo — collected operating-system and system information.

  • nltest /dclist: — queried domain controller information.

  • whoami /groups — identified the groups associated with the current security context.

  • net group domain admins /dom — enumerated members of the Domain Admins group.

These commands can be used legitimately by administrators, but their execution in suspicious sequences or from unexpected processes may indicate attacker reconnaissance.

Defenders should correlate command execution with process ancestry, user identity, endpoint activity, and network connections rather than treating any single command as conclusive evidence of compromise.

Privilege Escalation and Credential Access

The attackers created two accounts:

  • backup_DA

  • backup_EA

The backup_EA account was added to the Enterprise Administrators group, providing a significant privilege escalation opportunity within the Active Directory environment.

The attackers also accessed a domain controller and used wbadmin.exe to back up the following sensitive data:

  • NTDS.dit, which stores Active Directory directory information.

  • The SYSTEM registry hive.

  • The SECURITY registry hive.

The attackers additionally attempted to dump LSASS memory using rundll32.exe and comsvcs.dll, a technique that can expose credentials and authentication material.

They also attempted to extract credentials from a Veeam PostgreSQL database.

This activity demonstrates how attackers may target directory databases, registry hives, process memory, and backup infrastructure to expand their access and prepare for ransomware deployment.

Detection opportunities:

  • Alert on unexpected creation of privileged accounts.

  • Monitor additions to Domain Admins and Enterprise Administrators groups.

  • Investigate unusual use of wbadmin.exe on domain controllers.

  • Monitor suspicious execution of rundll32.exe involving comsvcs.dll.

  • Investigate unexpected access to LSASS memory and Veeam database files.

  • Review administrative logons and access to domain controllers from unusual endpoints.

Persistence and Lateral Movement

The attackers installed RustDesk on compromised hosts to establish remote access and support continued operations.

They also established an SSH reverse tunnel to:

193[.]242[.]184[.]150

A network scanner associated with the activity was renamed to n.exe, potentially making it less obvious during casual inspection.

The attackers continued investigating the environment and attempted to obtain credentials associated with Veeam infrastructure. The use of remote-access software and tunnelling can allow attackers to maintain connectivity even when their initial malware is removed.

Two days after the initial ransomware activity, the attackers reportedly returned through RustDesk, accessed a child domain controller, performed further discovery, and deployed Akira ransomware in the child domain.

Detection opportunities:

  • Investigate unauthorized RustDesk installations and unexpected remote sessions.

  • Monitor for new SSH tunnels and unusual outbound SSH connections.

  • Look for renamed administrative or network-scanning utilities.

  • Audit remote-access permissions and authentication events.

  • Investigate repeated access to domain controllers after an apparent incident has been contained.

Data Exfiltration

Before ransomware deployment, the attackers used FileZilla and SFTP to transfer data from a file server to external infrastructure.

The reported SFTP destination was:

185[.]174[.]100[.]203

The presence of legitimate file-transfer software does not itself establish malicious activity. However, an unexpected FileZilla installation or unusual SFTP transfers from a sensitive file server can indicate data theft when correlated with other suspicious behavior.

Organizations should monitor outbound file-transfer activity, investigate connections to unfamiliar destinations, and correlate large or unusual transfers with account activity and endpoint alerts.

Akira Ransomware Deployment

The attackers ultimately deployed the Akira ransomware payload named locker.exe.

The reported SHA-256 hash was:

de730d969854c3697fd0e0803826b4222f3a14efe47e4c60ed749fff6edce19d

The ransomware affected local and remote shares. By this stage, the attackers had already performed reconnaissance, established remote access, targeted privileged accounts, and attempted to access sensitive credentials and backup infrastructure.

The report described a time to ransomware of approximately 44 hours for the principal incident. It also referenced a separate Swisscom B2B CSIRT incident with a reported time to ransomware of approximately nine hours.

The Swisscom investigation contained related indicators and tooling, including a malicious Advanced-IP-Scanner.msi installer, a malicious DLL, an AdaptixC2 endpoint, and an SSH tunnel endpoint. These indicators should be treated as belonging to the related investigation rather than automatically attributed to the primary incident.

The case demonstrates that ransomware deployment is often the final stage of a broader intrusion, not the beginning of the attack. Detecting credential theft, privileged account changes, remote-access installation, and suspicious data transfers can provide opportunities to intervene before encryption.

Detection Engineering and Threat Hunting

Defenders should investigate the incident as a sequence of related behaviors rather than depending on a single malware signature or IOC.

1. Windows account and group changes

Review Windows Security logs for account creation and privileged group membership changes. Relevant events may include:

  • Event ID 4720 — user account created.

  • Event ID 4732 — member added to a security-enabled local group.

  • Event ID 4672 — special privileges assigned to a new logon.

These event IDs are investigation starting points. Validate the actual account, group, host, and event details before determining whether an action is malicious.

2. Suspicious process execution

Investigate process creation involving:

  • consent.exe loading an unexpected DLL.

  • rundll32.exe executing with comsvcs.dll.

  • Unexpected wbadmin.exe execution on domain controllers.

  • Unusual execution of n.exe, particularly when associated with network discovery.

Event ID 4688 can support process-creation investigations when the relevant auditing is enabled. PowerShell Script Block Logging events, including Event ID 4104, may also be useful if suspicious PowerShell activity is observed.

3. Network and command-and-control activity

Hunt for outbound connections to the reported infrastructure:

  • 109[.]205[.]195[.]211:443

  • 188[.]40[.]187[.]145:443

  • 172[.]96[.]137[.]160:443

  • 193[.]242[.]184[.]150

  • 185[.]174[.]100[.]203

Investigate the associated process, initiating user, destination port, timing, and surrounding endpoint activity. IP matches alone do not prove compromise, and infrastructure may change over time.

4. Unauthorized remote access and file transfer

Review endpoint inventories, software installation logs, remote-access sessions, firewall logs, and proxy records for unexpected RustDesk, FileZilla, and SSH activity.

Pay particular attention to systems that also show privileged account changes, domain controller access, suspicious file operations, or unusual outbound transfers.

5. Backup and directory database access

Investigate unexpected backups or access involving NTDS.dit, the SYSTEM and SECURITY registry hives, LSASS memory, and Veeam database files. Validate whether the activity was initiated by authorized administrators and whether the timing and destination are consistent with normal operations.

Indicators of Compromise (IOCs)

The following indicators were reported in the primary investigation and the related Swisscom investigation. Domains and IP addresses are defanged for safe handling. Validate them against trusted reporting before using them in production detection rules.

Domains and websites

  • opmanager[.]pro — malicious website associated with the trojanized ManageEngine OpManager installer.

  • ev2sirbd269o5j[.]org — reported Bumblebee domain-generation algorithm (DGA) domain.

  • 2rxyt9urhq0bgj[.]org — reported Bumblebee DGA domain.

  • angryipscanner[.]org — related malicious software-download website.

  • axiscamerastation[.]org — related malicious software-download website.

  • ip-scanner[.]org — website associated with the related Swisscom investigation.

IP addresses and network endpoints

Primary investigation:

  • 109[.]205[.]195[.]211:443 — Bumblebee command-and-control endpoint.

  • 188[.]40[.]187[.]145:443 — Bumblebee command-and-control endpoint.

  • 172[.]96[.]137[.]160:443 — AdaptixC2 endpoint.

  • 193[.]242[.]184[.]150 — SSH reverse-tunnel endpoint.

  • 185[.]174[.]100[.]203 — reported SFTP data-transfer destination.

Related Swisscom investigation:

  • 170[.]130[.]55[.]223 — AdaptixC2 endpoint.

  • 83[.]229[.]17[.]60 — SSH tunnel endpoint.

Filenames and artifacts

  • ManageEngine-OpManager.msi — trojanized installer.

  • msimg32.dll — malicious DLL associated with Bumblebee delivery.

  • AdgNsy.exe — AdaptixC2 beacon.

  • n.exe — renamed network-scanning utility.

  • locker.exe — Akira ransomware payload.

  • Advanced-IP-Scanner.msi — installer associated with the related Swisscom investigation.

  • win.exe — ransomware-related artifact associated with the related Swisscom investigation.

SHA-256 hashes

Primary investigation:

  • ManageEngine-OpManager.msi
    186b26df63df3b7334043b47659cba4185c948629d857d47452cc1936f0aa5da

  • msimg32.dll
    a6df0b49a5ef9ffd6513bfe061fb60f6d2941a440038e2de8a7aeb1914945331

  • locker.exe
    de730d969854c3697fd0e0803826b4222f3a14efe47e4c60ed749fff6edce19d

Related Swisscom investigation:

  • Advanced-IP-Scanner.msi
    a14506c6fb92a5af88a6a44d273edafe10d69ee3d85c8b2a7ac458a22edf68d2

  • msimg32.dll
    6ba5d96e52734cbb9246bcc3decf127f780d48fa11587a1a44880c1f04404d23

  • win.exe
    18b8e6762afd29a09becae283083c74a19fc09db1f2c3412c42f1b0178bc122a

These indicators represent different stages and investigations. Their presence alone should not be treated as conclusive proof of compromise, and the absence of a listed indicator does not rule out an intrusion.

Prevention and Security Lessons

The Bumblebee-to-Akira intrusion demonstrates how attackers can combine a malicious software download, command-and-control frameworks, credential theft, remote access, and data exfiltration to prepare for ransomware deployment.

Organizations should adopt the following safeguards:

  • Download software only from verified vendor websites and trusted distribution channels.

  • Use application control and endpoint detection to investigate suspicious installers and DLL-loading behavior.

  • Monitor the creation of privileged accounts and changes to sensitive Active Directory groups.

  • Restrict administrative access to domain controllers and separate privileged accounts from routine user accounts.

  • Enable appropriate Windows process-creation and security auditing.

  • Detect unauthorized remote-access software, SSH tunnels, and unexpected SFTP transfers.

  • Protect LSASS, Active Directory databases, and backup infrastructure with strict access controls.

  • Monitor Veeam systems and investigate unexpected database access or credential extraction attempts.

  • Maintain offline or immutable backups and regularly test recovery procedures.

  • Segment networks to limit lateral movement between workstations, servers, and domain controllers.

  • Investigate suspicious activity across the entire attack chain before returning affected systems to service.

Conclusion

The Bumblebee and Akira incident illustrates how search-engine poisoning and a trojanized software installer can lead to a serious enterprise compromise. After the initial Bumblebee execution, the attackers deployed AdaptixC2, conducted reconnaissance, created privileged accounts, attempted credential theft, established remote access, exfiltrated data, and deployed Akira ransomware.

The subsequent return through RustDesk and deployment in a child domain further demonstrate why incident response must address persistence and potential reinfection, not just the original ransomware payload.

For defenders, the central lesson is to detect and disrupt attacker behavior early. Monitoring suspicious software execution, privileged account changes, credential access, remote-access tools, and unusual data transfers can help prevent an initial compromise from escalating into a ransomware incident.

MITRE ATT&CK Mapping

T1189 — Drive-by CompromiseT1105 — Ingress Tool TransferT1059.003 — Windows Command ShellT1087.002 — Domain Account DiscoveryT1069.002 — Domain Groups DiscoveryT1098 — Account ManipulationT1003.001 — LSASS MemoryT1003.003 — NTDST1219 — Remote Access SoftwareT1572 — Protocol TunnelingT1041 — Exfiltration Over C2 ChannelT1567 — Exfiltration Over Web ServiceT1486 — Data Encrypted for Impact