Skip to main content

Hit by ransomware? Isolate affected systems now. Do not reboot or reformat.

SheMo Noransom舍末无勒

Ransomware family

Interlock Ransomware Decryption & Data Recovery

  • Active
  • High
  • No public decryptor

Interlock emerged in September 2024 as a double-extortion crew marked by .interlock / .1nt3rlock extensions and the !__README__!.txt note. It relies on drive-by downloads, ClickFix fake-CAPTCHA social engineering and edge-device zero-days, and was covered by a CISA #StopRansomware advisory in 2025. No public decryptor exists.

First seen
2024-09
File extensions
.interlock .1nt3rlock
Ransom notes
!__README__!.txt
Affected platforms
Windows / Linux / VMware ESXi

Family profile

File extensions
  • .interlock
  • .1nt3rlock
Ransom notes
  • !__README__!.txt
Contact patterns
  • Tor (.onion) negotiation portal + per-victim code
  • Interlock - Worldwide Secrets Blog leak site on Tor
  • No mailbox; the amount is quoted only after the victim makes contact
Aliases / versions
Interlock Ransomware、Interlock RAT / NodeSnake 关联团伙、Interlock FreeBSD Encryptor、Hive0163(IBM X-Force 追踪编号)
First seen
2024-09
Status
Active
Threat level
High
Affected platforms
  • Windows
  • Linux
  • VMware ESXi
Tags
  • Active
  • Leak-site regular
  • Double extortion
  • Targets virtualization
  • Phishing
  • Exploits vulnerabilities
Decryptor
No public decryptor

No free public Interlock decryptor exists. The encryptor combines AES and RSA, wrapping a per-file symmetric key with an RSA public key, and no exploitable implementation flaw has been published. The CISA/FBI #StopRansomware advisory AA25-203A of July 2025 offers detection and mitigation guidance but no decryption path. Tools advertised online as Interlock decryptors are untrustworthy. Recovery must rely on backups and snapshots, structural repair of virtual disks and databases, unencrypted duplicates and carving.

Latest activity

  1. Symantec detailed Mistic, an in-memory backdoor used since April 2026 by access broker KongTuke/Woodgnat, delivered through MpExtMs.exe DLL sideloading. KongTuke sells network access to Interlock, Rhysida, Akira and others.

    Sources
  2. Amazon threat intelligence disclosed that Interlock exploited CVE-2026-20131 in Cisco Secure Firewall Management Center as a zero-day from 26 January, 36 days before Cisco's 4 March advisory, for unauthenticated root code execution.

    Sources
  3. IBM X-Force reported Slopoly, a PowerShell backdoor likely written with generative AI, in an Interlock intrusion: ClickFix entry, a Runtime Broker scheduled task, and over a week of dwell time before exfiltration and encryption.

    Sources

Overview

Interlock was first identified in late September 2024. Its operating model differs from mainstream RaaS - no large-scale affiliate recruitment appears in public reporting, and it reads more like a closed, compact team - yet its tradecraft is distinctive and its reach into critical infrastructure is significant. CISA, the FBI, HHS and MS-ISAC published the joint advisory #StopRansomware: Interlock (AA25-203A) on 22 July 2025, marking its arrival on the regulatory radar.

The initial access methods stand out most. The FBI observed drive-by downloads from compromised legitimate websites, with payloads disguised as browser or security software updates - unusual among ransomware crews. The second main line is ClickFix social engineering: a fake CAPTCHA page persuades the user to copy and run a PowerShell command themselves, later joined by a FileFix variant using the file manager address bar. Both routes bypass traditional email attachment inspection entirely.

In 2026 the group escalated further, exploiting a critical Cisco Secure Firewall Management Center flaw, CVE-2026-20131, as a zero-day more than a month before public disclosure; Amazon's threat intelligence team issued a warning about the campaign. As of September 2026 its leak site listed 124 claimed victims - not the largest volume, but high-value targets concentrated in healthcare, manufacturing, technology, government and critical infrastructure, mainly across North America and Europe.

IBM X-Force reports overlap between Interlock and Rhysida in the Supper backdoor, crypters and the TAG-124/KongTuke traffic distribution system, suggesting a relationship in personnel or resources. For Chinese organisations, ClickFix-style attacks that get the user to paste and execute commands themselves have already appeared in local variants, and combined with fake software-update pages they are a priority concern on the endpoint. No public report confirms a targeted Interlock campaign against mainland China organisations.

How to identify it

Extension. .interlock or .1nt3rlock (leetspeak) is appended, with the original filename preserved.

Ransom note. !README!.txt, whose leading punctuation usually sorts it to the top of directory listings. The CISA advisory notes it is frequently delivered through a group policy object, which itself indicates the domain controller is under attacker control. The note supplies a victim-specific code and an onion address but no email and no figure - the victim must initiate contact through Tor first.

Intrusion-side indicators, often visible earlier than encryption.

  • Users reporting that a website prompted a "browser update" and delivered an installer.
  • Fake CAPTCHA pages instructing users to press Win+R and paste a command (ClickFix), or to run it via the file manager address bar (FileFix).
  • A 64-bit executable on endpoints masquerading as conhost.exe.
  • Interlock RAT, NodeSnake RAT, Cobalt Strike or SystemBC command-and-control components.
  • Credential stealers including cht.exe, Lumma Stealer and Berserk Stealer.
  • Azure Storage Explorer and AzCopy appearing on non-administrative hosts, indicating exfiltration.
  • AnyDesk, PuTTY and RDP used for lateral movement.

Self-deletion. After encryption a DLL such as tmp41.wasd invokes a remove routine that deletes the encryptor itself, so forensic work should include recovery of deleted files.

Infection vectors

Interlock's chain runs along two main lines - getting the user to act, and hitting edge devices:

  • Drive-by downloads. Compromised legitimate websites serve visitors a malicious installer posing as a browser or security software update. Because the source is a site the user visits routinely, success rates are high.
  • ClickFix / FileFix social engineering. A fake human-verification page walks the user through running a PowerShell command themselves, or executing it via the file manager address bar, bypassing email and download inspection.
  • Edge device zero-days and known flaws. In 2026 CVE-2026-20131 in Cisco Secure FMC was exploited as a zero-day against the enterprise firewall management plane.
  • Traffic distribution systems. Payload delivery through TDS infrastructure such as TAG-124/KongTuke.
  • Persistence and C2. Interlock RAT, NodeSnake RAT, Cobalt Strike and SystemBC.
  • Credential theft and lateral movement. Multiple infostealers harvest credentials, then AnyDesk, PuTTY and RDP carry the operators toward domain controllers and the virtualisation platform.
  • Exfiltration. Azure Storage Explorer and AzCopy upload data to cloud storage, so the traffic resembles ordinary cloud sync and is harder to detect.
  • Mass deployment. Encryptor and note pushed domain-wide via GPO.

ClickFix-style attacks work by getting the user to execute code manually, past every technical control. Endpoint hardening therefore has to be paired with awareness training and one unambiguous rule: any page that asks you to copy and paste a command is an attack.

Encryption behavior

Algorithms. The encryptor combines AES and RSA: each file's content is encrypted with its own AES key, which is then wrapped with an RSA public key and written into the file. Without the operators' private key nothing can be reversed, and the CISA advisory discloses no exploitable implementation flaw.

Multi-platform coverage. The Windows encryptor is a 64-bit executable, often masquerading as conhost.exe. A Linux ELF encryptor also exists, and unusually so does a FreeBSD ELF encryptor, analysed as compiled on FreeBSD 10.4. Public analysis holds that these non-Windows builds are aimed primarily at virtual machine image files on virtualisation and storage platforms; before Interlock, only the now-dismantled Hive operation had produced a FreeBSD encryptor.

Execution and clean-up. After encryption a helper DLL deletes the encryptor itself to reduce forensic traces, alongside event log clearing and weakening of endpoint protection.

Recovery inhibition. Volume shadow copies deleted, database and backup services stopped, and network shares and reachable backup locations encrypted where possible.

Double extortion. Data is exfiltrated to cloud storage via Azure Storage Explorer and AzCopy before encryption and published on the Interlock leak site if unpaid. The note deliberately omits any figure and requires the victim to make contact first - a negotiation tactic that allows pricing to be set against the victim's size.

Assess before you act

Recoverability assessment

There is no public decryptor for Interlock and decryption is not technically feasible. We do not pay ransoms and do not negotiate.

1) Backups, snapshots and shadow copies (the most important path). Shadow copies are usually gone, but offline media, backup appliances with independent credentials, offsite copies, storage array and NAS volume snapshots, hypervisor snapshots and cloud version history frequently survive. Because Interlock commonly pushes domain-wide via GPO, every domain-joined backup system must be assumed reachable - prioritise backups authenticating outside the domain and verify usability by actual mount.

2) Structural repair of VM and database files (depends on the encryption pattern). For large files, encryptors trading against time rarely overwrite every byte, so vmdk, vhdx and database files may retain usable intact regions. We measure the coverage first, then decide whether to repair partitions and metadata for mounted extraction, or to run page-level extraction and logical rebuilds on SQL Server, Oracle and MySQL with transaction log roll-forward. Yields depend on the damage to critical structures.

3) Unencrypted duplicates and log replay. File-server previous versions and recycle bins, endpoint caches, reporting and BI staging databases, ERP and HIS archive exports, mail copies and application audit logs are realistic sources for rebuilding critical records.

4) Low-level carving. Interlock deletes its own encryptor after execution, and in some cases original files are also written new and deleted, leaving intact content in unallocated space that raw sector extraction can recover - provided writes to the affected volumes stop immediately.

5) Cloud-side traces have double value. Interlock exfiltrates using AzCopy and similar tools, leaving logs on both the cloud storage and endpoint sides. These are required for scoping the leak for compliance purposes, and sometimes also reveal which data already had a cloud copy before encryption, opening an additional recovery source.

For Interlock data recovery we provide a measured assessment and a bounded scope first. We do not claim "100% decryption" and no approach guarantees full recovery.

Our response plan

Hit by Interlock ransomware? What to do

  1. Containment and forensic preservation

    Isolate affected hosts, hypervisors and domain controllers, break storage paths, and revoke active sessions and cloud storage access keys - Interlock exfiltrates with AzCopy and those keys may still be valid. Do not reboot and do not re-initialise datastores. Image or take read-only snapshots of domain controllers, backup servers and datastores, and export Active Directory and GPO change records, proxy and DNS logs, cloud storage access logs and EDR telemetry. Retain three to five encrypted files and !README!.txt, and because the encryptor self-deletes, run deleted-file recovery on the system volume to retrieve the sample.

  2. Family identification and encryption analysis

    Confirm the family from the .interlock / .1nt3rlock extension, the !README!.txt format and encrypted-file structure, and separate the Windows, Linux and FreeBSD builds. Trace whether initial access came from a drive-by download, ClickFix/FileFix social engineering or an edge-device vulnerability - this directly shapes the cleanup scope and hardening priorities. In parallel, measure the encrypted proportion of large files and map intact data inside virtual disks and databases.

  3. Recoverability assessment and plan sign-off

    Check backup usability in layers - backups authenticating outside the domain, then storage and hypervisor snapshots, then domain-joined backup systems - verifying each by actual mount. Run sample repairs on core databases and key VMs to quantify what is recoverable, and in parallel scope the exfiltration from cloud storage access logs and endpoint telemetry. Deliver a written assessment covering recovery path, expected scope, timeline and business sequencing per system, then execute after sign-off.

  4. Recovery execution

    Work on images in a clean environment with originals read-only. In virtualised estates, repair virtual disk structures and extract via mount, landing data on new storage before rebuilding VMs. Databases go through page-level extraction plus transaction log roll-forward, and filesystem data is restored in batches by business directory, supplemented by deleted-file recovery where needed. Sequence: identity and network infrastructure, then core business systems, then collaboration and archives, with hash verification, application open tests and business reconciliation after each batch.

  5. Attribution, hardening and handover

    Reconstruct the full kill chain and harden against Interlock's specifics: on endpoints, prevent ordinary users from executing pasted PowerShell (execution policy plus Win+R restrictions), deploy browser download controls and software installation allowlisting, and add detection for fake update packages; on the network, patch and reduce exposure of management planes such as Cisco FMC; remove Interlock RAT, NodeSnake, Cobalt Strike, SystemBC and unauthorised remote control tools; reset domain credentials and cloud storage keys; put GPO changes under audit alerting; and rebuild backups around out-of-domain authentication, immutable copies and offline media. Close with an incident report, exfiltration conclusion, handover checklist and a staff exercise focused on ClickFix-style social engineering.

Risk warning

What not to do

  • Do not follow any web page that tells you to copy a command and run it with Win+R or paste it into a file manager address bar - that is the ClickFix / FileFix attack itself, and genuine CAPTCHAs never ask users to execute commands.
  • Do not reboot or reinstall affected hosts. The Interlock encryptor self-deletes, and reinstalling or heavy disk writes destroy both the sample and the chance of deleted-file recovery.
  • Do not declare the incident closed after cleaning endpoints. Interlock commonly deploys domain-wide via GPO, so domain controllers and group policy must be fully audited or the payload simply returns.
  • Do not overlook the cloud side: access keys and tokens used for exfiltration must be revoked immediately and the associated access logs exported and preserved as the key evidence for scoping the leak.
  • Do not connect to backup systems with domain credentials, and do not resume operations before remote control and C2 components are removed - repeat encryption is common in cases of this family.
  • Do not contact the onion portal or pay. The note deliberately omits a figure so the operators can price against your size after contact; payment guarantees neither decryption nor non-publication.

Emergency response

Data already encrypted? Stop and let an engineer look first

We do not pay ransoms and we do not negotiate with attackers. Engineers run a free assessment first, then propose a recovery plan and a firm quote.

Related scenarios

Related industries

Similar families

FAQ

Interlock Frequently asked questions

  • Is there a decryptor for .interlock files?

    No. Interlock encrypts content with AES and wraps the key with RSA, and the CISA/FBI #StopRansomware advisory AA25-203A of July 2025 offers only detection and mitigation guidance, no decryption path. Anything online claiming to decrypt .interlock or .1nt3rlock should be treated as malware. Real recovery runs through backups and snapshots, structural repair of virtual disks and databases, unencrypted duplicates, and low-level extraction of deleted files.

  • Staff say they clicked a browser update on a normal website - is that plausible?

    Entirely plausible - it is one of Interlock's signature entry points. The FBI observed the crew compromising legitimate websites and serving visitors installers disguised as browser or security software updates; because the source is a site the user visits routinely, suspicion is low. The other main line is ClickFix: a fake human-verification page persuades the user to press Win+R and paste a PowerShell command, later joined by a FileFix variant using the file manager address bar. Both bypass email attachment inspection and rely on the user executing the code themselves.

  • The ransom note does not state an amount - why?

    It is a deliberate negotiation design. The note supplies only a victim-specific code and an onion address - no email and no figure - requiring the victim to make contact through Tor first, after which pricing is set against their size, sector and data sensitivity. Operationally this means that simply making contact to see the number already hands the operators information and raises the quote. Our advice is not to engage, and to spend the effort on scoping the exfiltration, verifying backups and driving technical recovery.

  • Are Interlock and Rhysida the same people?

    Not settled, but the overlap is clear. IBM X-Force documents shared use of the Supper backdoor, crypters and TDS infrastructure such as TAG-124/KongTuke, and assesses a probable relationship in personnel or resources; Rhysida dates from 2023 and Interlock from September 2024. For victims the practical value is that detection and hardening transfer: indicators and detection rules built for Rhysida often catch Interlock, and vice versa.

  • A zero-day in our firewall management platform was exploited - how do we defend against that?

    A zero-day cannot be patched in advance, but the blast radius can be cut sharply. Four measures matter most: never expose management planes - firewall management centres, virtualisation management, bastion hosts - to the internet, keeping them in a dedicated segment reachable only from specific jump hosts; keep management platform accounts separate from the production domain with MFA and source IP restrictions; build audit alerting for configuration changes, anomalous logins and new accounts on those platforms; and maintain a vendor advisory channel with a pre-agreed emergency upgrade process. In the 2026 Cisco FMC cases, how quickly anomalous management activity was noticed made the difference in impact.