Ukrainian police say a ransomware organization entered company networks by impersonating IT support. Employees were persuaded to disclose credentials or install remote-access software. An unexpected support call should be verified through a known helpdesk before it becomes permission to enter a device.
The November 18 police announcement describes a joint investigation with the FBI lasting more than two years. Investigators say at least 28 companies paid more than $40.5 million. Four suspects are in custody; the investigation continues. These are police findings and allegations, not convictions or a complete measure of all victims’ losses.
The caller kept talking while another attacker logged in
The police describe one documented episode in which an employee remained on the phone with a purported support specialist. A second attacker entered the corporate system. The call ended only after the attackers confirmed access.
Investigators also report a platform using AI-generated voice messages to automate contacts with employees. Their account does not establish that every call used a cloned voice or that AI was necessary for the intrusion. The decisive access came from credentials or remote-control software supplied by the employee.
Lawrence Abrams’s reporting adds details attributed to Ukrainian court records. These include fake corporate VPN login pages, spoofed phone numbers, and remote-access tools such as TeamViewer and AnyDesk. In the reported Warren Oil case, the attackers allegedly obtained credentials and then a two-factor authentication code.
A second login factor can still be relayed by someone impersonating support. That case concerns a disclosed code; it is not evidence that every form of multifactor authentication was bypassed.
Verify the request separately from the caller
The exposure here is a workflow: an employee can supply access after receiving an apparently helpful call. The public sources identify no single vulnerable software version or patch that closes that route.
These cases support several practical controls:
- End an unexpected call and contact the known helpdesk yourself. Use an established directory or support portal, not a number or link supplied by the caller.
- Keep login credentials and one-time codes out of support conversations. A caller’s knowledge of the company, employee or VPN service does not establish their authority.
- Require an independently verified support request before installing or enabling remote-control software. An authentic product can still give an impostor access.
- Give employees a quick way to report a suspicious request, including after complying. The response team can revoke sessions, disable unauthorized remote access and preserve relevant evidence.
- Review identity and remote-access records together after a suspected intrusion. A familiar support tool’s presence alone cannot distinguish authorized help from attacker control.
Police say the organization stole confidential data and encrypted files, then threatened publication while demanding payment. Their example of a US company includes 1.4 terabytes stolen and an attack extended to other companies in its corporate group. The access review therefore needs to consider connected organizations as well as the first account or device.
Evidence status: real intrusions and arrests reported by investigators. BleepingComputer adds a court-record account; it does not independently establish every allegation. Police associate the wider network with former Conti members, but the arrested suspects’ individual ransomware affiliations remain unconfirmed.
Caveat: The public accounts provide selected cases, not a complete victim list or a demonstrated rate at which AI-generated calls succeed.