Ransomware family
Lynx Ransomware Decryption & Data Recovery
- Active
- High
- No public decryptor
Lynx is a RaaS operation that emerged in mid-2024 and shares extensive code with INC Ransom. It ships encryptors for Windows, Linux and ESXi through an affiliate panel with an 80/20 split and had passed 400 claimed victims by 2026. No public decryptor exists.
- First seen
- 2024-07
- File extensions
- .LYNX .lynx
- Ransom notes
- README.txt
- Affected platforms
- Windows / Linux / VMware ESXi
Family profile
- File extensions
- .LYNX
- .lynx
- Ransom notes
- README.txt
- Contact patterns
- Tor (.onion) negotiation portal (several mirror addresses)
- Chat system built into the affiliate panel
- Lynx Blog leak site on Tor
- Ransom notes printed to internal network printers
- Aliases / versions
- Lynx Ransomware、INC Ransom 代码衍生品牌、Lynx ESXi Locker
- First seen
- 2024-07
- Status
- Active
- Threat level
- High
- Affected platforms
- Windows
- Linux
- VMware ESXi
- Tags
- Leak-site regular
- Active
- Ransomware-as-a-Service
- Double extortion
- Targets virtualization
- RDP brute force
- Phishing
No free public Lynx decryptor exists. The encryptor performs key agreement with Curve25519 Donna and encrypts content with AES-128-CTR using a per-file key, with no published implementation flaw. Lynx shares extensive code with INC Ransom (roughly 48% function overlap, 70.8% among common functions) but their key systems are independent, so no universal solution exists and tools claiming to decrypt .LYNX files are untrustworthy. Recovery runs through backups and snapshots, structural repair of virtual disks and databases, unencrypted duplicates and carving.
Latest activity
Lynx's newest leak-site entry is US investment adviser Cutler Capital Management, lifting its total to ~418. Posting slowed to sporadic batches in 2026 and most onion mirrors show ~56% uptime, yet Lynx is not defunct.
SourcesOn 10 May 2026 Lynx batch-posted victims including Taiwanese battery maker CSB Energy Technology and Germany's Kurita Europe, showing its self-declared 'no China' rule does not cover Greater China or overseas units.
Sources
Overview
Lynx appeared in July 2024 and began recruiting affiliates on the RAMP forum in August under the handle "silencer", quickly building a scaled ransomware-as-a-service operation. Unit 42's binary comparison established roughly 48% overall function overlap with INC Ransom, rising to 70.8% among shared functions and higher still in the Linux/ESXi variant. Combined with the public sale of INC's source code in 2024, the consensus view is that Lynx is a repackaging of purchased or reused INC code.
The operation is notably productised. Group-IB's analysis of the affiliate panel describes modules for News, Companies, Chats, Stuffers and Leaks, supporting victim profile management, custom sample generation, negotiation chat and scheduled leak publication, with pre-built binaries covering Windows, Linux across x86, ARM, MIPS and PPC, and ESXi. The affiliate split is 80/20 in the affiliate's favour. That level of industrialisation lets crews without development capability attack virtualisation estates efficiently.
Victims were initially concentrated among US and UK retail, real estate, architecture, financial and environmental services firms. From early 2026 Lynx began posting in large single-day batches - 20 organisations on 5 January alone - passing 400 claimed victims in total, placing it in the upper-middle tier of active families.
One point matters specifically for Chinese organisations: Lynx's recruitment posts explicitly state that it does not operate in the CIS, Ukraine, China, Iran or North Korea, and does not target healthcare, government institutions, churches or children's charities. Such statements are not a security guarantee - affiliates do not always comply, and overseas entities and cross-border systems of Chinese groups fall outside that self-declared restriction - but there is currently no public report of a Lynx campaign targeting mainland China organisations.
How to identify it
Extension. .LYNX is appended to encrypted files (lowercase .lynx in some builds), with the original filename preserved.
Ransom note. A README.txt appears in affected directories; its body is stored Base64-encoded in the sample and decoded at runtime. The English text states that files are encrypted and data stolen, and supplies several Tor mirror addresses plus victim-specific login credentials.
Distinctive behaviour.
- Desktop wallpaper replaced with a ransom message.
- The note printed in bulk to reachable network printers; some early samples attempted to print a report through Microsoft OneNote.
- In ESXi environments, virtual machines force-stopped and snapshots deleted.
Telling Lynx from INC. The encryption flow is similar, but the extension (.LYNX versus .INC), note filename (README.txt versus INC-README.txt/html), leak site and key systems all differ. Because the code is shared, behaviour alone is easy to confuse, so accurate attribution requires comparing encrypted-file trailer structures and binary fingerprints.
Other indicators. Anomalous RDP logins before the attack, installation of remote control tools such as AnyDesk, credential dumping utilities, and archiving or cloud upload tools appearing on non-administrative hosts.
Infection vectors
Lynx operates as RaaS, so specific tradecraft varies by affiliate, but public cases repeatedly show:
- Valid credentials plus remote access. Purchased credentials, infostealer logs or password attacks used to log into RDP, VPN and remote maintenance entry points - the dominant starting position.
- Phishing. Loader delivery to establish a foothold before pulling down further tooling.
- Edge and application exploitation. Unpatched internet-facing services used for entry.
- Lateral movement. RDP, PsExec, WMI and remote control tools such as AnyDesk combined with stolen domain accounts.
- Escalation and defence evasion. Credential dumping for domain admin rights, disabling or uninstalling endpoint protection, clearing logs.
- Hypervisor takeover. Once vCenter or ESXi credentials are obtained, the Linux/ESXi encryptor is deployed at host level, taking down every VM at once.
- Exfiltration completed before encryption as leak leverage.
Because the affiliate panel ships binaries for many architectures, Lynx affiliates have a wider blast radius in mixed estates - Windows domains plus Linux servers plus ESXi clusters plus assorted embedded devices - than most families. That is particularly unfavourable for the ageing-Linux-servers-and-virtualisation-cluster combination common in Chinese manufacturing.
Encryption behavior
Algorithms. The Windows encryptor performs Curve25519 Donna key agreement and encrypts content with AES-128-CTR using a per-file key; the Linux/ESXi variant uses the same cryptographic design.
Encryption modes. Four configurable levels are offered: fast (about 5%), medium (about 15%), slow (about 25%) and entire (100%). This matters enormously for recovery assessment. If an affiliate chose fast or medium, 75% to 95% of the blocks in large files - databases, virtual disks, backup containers - are completely untouched and structural repair has substantial room. If entire was chosen, there is effectively none. The encrypted proportion must therefore be measured before a recovery path is chosen; no general rule applies.
Command-line control. Target directories, single files, exclusion paths and hidden execution are all selectable, letting affiliates tune behaviour to the environment.
Virtualisation handling. The Linux/ESXi build force-terminates running virtual machines and deletes VM snapshots before encrypting files under /vmfs/volumes, ensuring complete encryption and making rollback difficult.
Recovery inhibition. The Windows build deletes volume shadow copies through DeviceIoControl calls and attempts to damage backup partition drives.
Double extortion. Data is exfiltrated before encryption and published in stages if unpaid, with wallpaper and printed notes adding pressure channels.
Assess before you act
Recoverability assessment
There is no public decryptor for Lynx and decryption is not viable. We do not pay ransoms and do not negotiate. What Lynx does offer is a clear handle for assessment, thanks to its configurable modes.
1) Measure the encrypted proportion first - it determines everything else. Because Lynx offers fast, medium, slow and entire modes, two incidents with the same .LYNX extension can differ enormously in actual damage. We measure coverage across sample files of different types and sizes: at fast (about 5%) or medium (about 15%), the repairable share of large database files and virtual disks is usually substantial; at entire, recovery must pivot entirely to backups.
2) Backups, snapshots and shadow copies. Lynx deletes shadow copies, attempts to damage backup partitions and removes VM snapshots on ESXi. But storage array and NAS volume snapshots, backup appliances with independent credentials, offline tape, offsite copies and cloud version history are usually beyond its reach and should be inventoried first and verified by actual mount.
3) Structural repair of virtual disks and databases (depends on the encryption pattern). In fast or medium mode, vmdk images typically retain long intact stretches of guest filesystem metadata and data, so partitions and metadata can be repaired and inner files extracted via mount. SQL Server, Oracle and MySQL data files support page-level extraction and logical rebuilds with transaction log roll-forward.
4) Unencrypted duplicates and log replay. File-server previous versions and recycle bins, endpoint caches, BI and reporting staging databases, ERP and MES archive exports, mail copies and application audit logs are standard sources for rebuilding critical records.
5) Low-level carving. If originals were deleted rather than overwritten in place, intact copies may survive in unallocated space and can be carved - provided writes to the affected volumes and datastores stop immediately.
Our approach to Lynx data recovery is to let measurement lead, then state a bounded scope and expectation. We do not promise "100% decryption" or guaranteed recovery.
Our response plan
Hit by Lynx ransomware? What to do
Containment and forensic preservation
Isolate affected Windows hosts, Linux servers and ESXi hosts, break storage paths, disable the remote access accounts involved and revoke active sessions. Do not reboot and do not re-initialise datastores. Image or take read-only snapshots of domain controllers, backup servers, vCenter and datastores, and export Active Directory, VPN, remote tooling and ESXi logs. Retain three to five .LYNX files, README.txt and the physical printed notes - their print timestamps help pin down the deployment window.
Family identification and encryption-mode measurement
Confirm Lynx from the extension, the Base64 body structure of README.txt and encrypted-file trailer fingerprints, distinguishing it from the code-related INC Ransom. Then measure the encrypted proportion across file types and sizes to determine whether the affiliate used fast, medium, slow or entire mode. This single determination decides whether recovery goes the structural repair route or depends wholly on backups, and is the most consequential judgement in handling this family.
Recoverability assessment and plan sign-off
Combine the measured encryption mode with an inventory of backup and snapshot usability - verified by actual mount rather than job status - and run sample repairs on core databases and key VMs to quantify what is recoverable. Deliver a written assessment covering, per business system, whether recovery uses backup rollback, structural repair or carving, with expected scope, timeline and priority, plus an initial exfiltration scope conclusion.
Recovery execution
Work on images in a clean environment with originals read-only. For ESXi, repair virtual disk structures and extract by mounting, landing data on new storage before rebuilding VMs. Databases go through page-level extraction plus transaction log roll-forward, and Linux server data is extracted in batches by directory. Sequence by business impact, and after each batch run hash verification, application open tests and business reconciliation into a traceable manifest.
Attribution, hardening and handover
Reconstruct the kill chain: where credentials leaked from (infostealer, password attack or phishing), when remote control tools were installed, how vCenter or ESXi credentials were obtained, and the timing and volume of exfiltration. Hardening: consolidate remote access behind MFA-protected entry points, remove unauthorised remote control software and enforce an allowlist, isolate the virtualisation management network with lockdown mode, redeploy EDR with tamper protection, rebuild backups around independent credentials, immutable copies and offline media, and run a dedicated inventory of multi-architecture Linux assets. Close with an incident report and handover checklist.
Risk warning
What not to do
- Do not conclude "everything is destroyed" or "it can all be repaired" before measuring the encrypted proportion - Lynx's four modes mean the same .LYNX extension can represent wildly different damage levels.
- Do not reboot ESXi hosts, re-initialise datastores or create new VMs on the original LUN; Lynx has already deleted VM snapshots, leaving low-level recovery as the only remaining path to original data.
- Do not download so-called Lynx decryptors. Lynx and INC share code but not keys, no universal solution exists, and such tools are typically secondary ransomware or infostealers.
- Do not discard README.txt, encrypted samples or the printed notes; print timestamps are valuable for reconstructing the deployment timeline.
- Do not use domain credentials to reach or rebuild backup systems; Lynx affiliates typically hold domain admin rights, and doing so can expose surviving backups.
- Do not resume operations before remote control tools such as AnyDesk and other persistence are removed, or a second encryption run becomes very likely.
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
ESXi / Hyper-V Virtualization Encrypted by Ransomware
Hypervisor-level encryption causes the widest blast radius of any ransomware event: dozens of production VMs go dark within an hour or two. This page covers what Linux ESXi encryptors actually do — shut down guests, encrypt vmdk, delete snapshots — the recovery value of flat disk files, and how Hyper-V and Proxmox cases differ.
File Servers and NAS Encrypted by Ransomware
When shared folders on a file server or NAS are encrypted, drawings, contracts, archives, quotations and design sources all become unusable at once — and mapped drives spread the impact to every endpoint. This page covers how to gauge spread, what shadow copies and snapshots realistically offer, and how to sequence recovery by business value.
Domain Controller Compromise and Estate-Wide Encryption
A compromised domain controller hands the attacker a legitimate administrator identity, allowing an encryptor to be pushed to every host at once through Group Policy or remote execution. This page covers how such incidents present, the correct order for Active Directory recovery, and how to decide between cleanup and full rebuild.
Backups Deleted or Destroyed
Modern ransomware follows a fixed sequence: destroy the backups, then encrypt the data — deleting shadow copies, encrypting repositories, disabling jobs, and exploiting backup software flaws to steal credentials. This page covers what can still be inventoried once backups fail, why replication propagates encrypted files off-site, and what offline and immutable copies are really worth.
Related industries
Manufacturing Ransomware Response and Recovery
Ransomware in manufacturing hits information systems and production cadence at the same time: with ERP down there are no orders, with MES down there is no schedule, and an encrypted drawing library takes the process documentation for an entire product line with it. This page covers the asset profile, recovery priorities and targeted defences.
Financial Services Ransomware Response and Recovery
Financial and quasi-financial institutions face far stricter requirements on data integrity, transaction continuity and regulatory reporting than most sectors, so one ransomware event hits availability, customer trust and compliance simultaneously. This page covers the threat profile, a recovery approach centred on transactional consistency, and hardening priorities.
Retail and E-commerce Ransomware Response
In retail and e-commerce, ransomware translates directly into an inability to sell: order systems, membership, POS and warehouse fulfilment stop together and losses accrue by the hour. This page covers the sector's attack patterns, a recovery order built around the order-to-fulfilment chain, and handling of member data exposure.
Similar families
- No public decryptor
INC Ransom
INC Ransom emerged in 2023 and has passed 800 claimed victims by 2026, making it a top-tier RaaS operation. Marked by the .INC extension and INC-README notes, it exploits Citrix and SonicWall edge flaws and ships an ESXi encryptor. No public decryptor exists.
- No public decryptor
SafePay
SafePay emerged in late 2024 and rose sharply through 2025-2026 as a closed, non-RaaS crew. Marked by the .safepay extension and readme_safepay.txt note, it enters mainly through valid credentials on VPN gateways and has passed 500 claimed victims. No public decryptor exists.
- No public decryptor
DragonForce
DragonForce is one of the most active ransomware cartels today. Since 2025 it has offered white-label encryptors and infrastructure to affiliates, hits virtualisation estates hard, and became widely known through the chain of UK retail attacks. No public decryptor exists.
FAQ
Lynx Frequently asked questions
Can .LYNX files be decrypted?
There is no public decryptor. Lynx uses Curve25519 Donna key agreement and AES-128-CTR content encryption with per-file keys and no known exploitable flaw. Although it shares extensive code with INC Ransom, their key systems are independent and no universal solution exists. Real Lynx data recovery starts by measuring the encrypted proportion, then combines backups and snapshots, structural repair of virtual disks and databases, unencrypted duplicates and carving.
Why can some Lynx victims repair data while others cannot?
Because the Lynx encryptor offers four modes: fast encrypts about 5%, medium about 15%, slow about 25% and entire 100%, with affiliates choosing based on their time window and target size. Under fast or medium, large database files and virtual disks retain many completely untouched blocks, so page-level extraction and structural repair succeed at a high rate. Under entire, files are fully overwritten and only backups remain. That is why our first step is always to measure coverage rather than judge from the extension.
Lynx says it does not target China - can Chinese firms relax?
No. Lynx's recruitment posts do state that it does not operate in the CIS, Ukraine, China, Iran or North Korea and avoids healthcare, government, churches and children's charities, and there is no public case report against mainland China organisations. But such declarations are an operational policy, not a guarantee: affiliates do not always comply, overseas subsidiaries, cross-border systems and joint ventures of Chinese groups fall outside that self-restriction, and the shared INC codebase carries no such limitation at all. The defensive work still needs doing.
Lynx encrypted our ESXi VMs and deleted the snapshots - what now?
Three things first: stop all writes to the datastore immediately - no new VMs, no expansion and absolutely no re-initialisation; check storage array or NAS volume snapshots and VM-level backups on the backup platform, neither of which is affected by snapshot deletion inside ESXi; and measure the encrypted proportion of the vmdk files. If the affiliate used fast or medium mode, structural repair of the virtual disks is viable - repairing partition tables and filesystem metadata then mounting to extract inner files often recovers more than expected.
We cannot tell Lynx from INC Ransom - does that affect the response?
Yes. They share code and behave similarly, but extensions, notes, leak sites and key systems all differ. Confusing them causes three problems: searching the wrong leak site during exfiltration investigation, applying recovery experience that does not fit, and getting external communications wrong. Technically the separators are the encrypted-file trailer structure, the note format (README.txt versus INC-README.txt/html) and binary fingerprint comparison. We settle the variant before designing the recovery plan.
Sources
- Cat's out of the bag: Lynx Ransomware-as-a-Service — Group-IB
- Lynx Ransomware: A Rebranding of INC Ransomware — Unit 42, Palo Alto Networks
- Ransomware Roundup - Lynx — FortiGuard Labs
- Dark Web Profile: Lynx Ransomware — SOCRadar
External links are provided for reference only. The content is published by third parties and does not represent our position.
Updated