Join our Newsletter — 33% off our NHI Course
Home› Guides› Rhadamanthys Infostealer Guide: Delivery, Data Theft and Detection…
Guide Attack Techniques & Malware

Rhadamanthys Infostealer Guide: Delivery, Data Theft and Detection for Defenders

← All guides
By NHI Mgmt Group, edited by Lalit Choda Updated 14 November 2026 15 min read
On this page

Rhadamanthys is a subscription infostealer that turns one opened attachment or pasted command into a package of browser data, stored credentials and cryptocurrency wallet files for whoever rented the build. It has been sold since 2022, has been delivered by several unrelated criminal groups, and was a named target of the November 2025 phase of Operation Endgame. This guide covers what the reporting vendors and Europol have published about the family, how its delivery chains look from the defender's side, which ATT&CK techniques apply, and how to detect, contain and respond to an infection. It is written for SOC analysts, endpoint and email security engineers and incident responders. For the wider stealer economy, read the Infostealer Threat Guide first; this page goes deeper on one family and does not repeat that material.

Key takeaways

  • It is a product, not a single actor. Check Point Research describes Rhadamanthys as malware-as-a-service sold on underground markets since 2022 and used by multiple actors, so lures, targets and infrastructure vary by customer.
  • Delivery is varied and user-driven. Published campaigns used invoice-themed password-protected ZIPs with LNK files, fake copyright notices linking to archives that side-load a DLL, and ClickFix paste-and-run lures.
  • Hash blocking alone fails. Builds are padded to change hashes, staged payloads are hidden inside image files, and file formats and mutex naming change between versions, so behavioural detection matters more than indicator lists.
  • The takedown did not expire the stolen data. Europol reported that the infrastructure taken down in November 2025 held several million stolen credentials; anything harvested before that date may still be traded or used.
  • Response is credential revocation. Treat every password, browser session, token and wallet secret reachable from an infected device as compromised, including the non-human credentials developers keep locally.

What Rhadamanthys is

Rhadamanthys is a modular information stealer for Windows. Check Point Research traced its first advertisement on an underground market to September 2022 and, in an August 2023 analysis, set out technical overlaps with an older malware project called Hidden Bee: custom executable formats, a virtual file system, reused functions, steganography and Lua scripting. Check Point's assessment is that Rhadamanthys continues the Hidden Bee work, likely by the same author or group, rather than simply borrowing from it. Its core modules use non-standard executable formats rather than normal Windows PE files, which makes memory scanning and analysis harder.

The family is sold as a subscription. In its October 2025 analysis of version 0.9.x, Check Point reported monthly tiers starting at 299 US dollars for a self-hosted panel and 499 US dollars with a rented server, plus a negotiated enterprise tier, and noted that the sellers had rebranded under the names "RHAD Security" and "Mythical Origin Labs" with a new website. That business model is why defenders see Rhadamanthys behind very different lures: each customer runs its own campaigns.

DateDevelopmentSource
September 2022First advertisement on an underground marketCheck Point Research, 2023
August 2023Analysis links custom executable formats to the Hidden Bee projectCheck Point Research
April 2024First observed use by TA547 against German organisations, invoice lures impersonating a retailerProofpoint
July to November 2024Version 0.7 in a large copyright-infringement phishing campaign across several regionsCheck Point Research
Early to late 2025Versions 0.9, 0.9.1 and 0.9.2; observed in ClickFix campaignsCheck Point Research
10 to 14 November 2025Operation Endgame phase takes down 1,025 servers across Rhadamanthys, VenomRAT and ElysiumEuropol, Operation Endgame

What it steals and why it is an identity problem

Check Point lists credentials, browser data, files and cryptocurrency wallets as the core targets of version 0.9.2, which added the Ledger Live wallet application and a browser fingerprinting module. Its 0.9.x analysis also names FTP, email and messaging clients, VPN software, two-factor authentication and password manager applications, and a long list of wallet applications. The 2023 analysis described modules aimed at wallet files, account profile files and KeePass credentials.

Version 0.7 added an optical character recognition module that the seller advertised as "AI-powered". Check Point found it used classic machine learning rather than a modern AI engine, and that it searches images and PDFs for words from the BIP-39 wallet recovery word list, which points at finding seed phrases people have photographed or scanned.

  • Human accounts: saved browser passwords, password manager data and email client credentials give direct access to corporate SaaS, webmail and VPN.
  • Browser sessions: the published reporting describes "browser data" without itemising cookies. Response should still assume that any browser session on the device is exposed, because session cookies replayed elsewhere can skip the MFA step already completed. See MITRE ATT&CK T1539 Steal Web Session Cookie.
  • Non-human credentials: on developer and administrator machines, files and application data can include cloud access keys, API tokens, SSH keys and CI/CD secrets. These do not expire on their own and are rarely rotated after an endpoint is reimaged. See the Infostealer Logs and Session Token Abuse Guide.
  • Crypto assets: wallet files and recovery phrases stored on disk or as images are a direct financial target. Europol said the main suspect behind the infostealer had access to over 100,000 crypto wallets.

Delivery chains seen in published campaigns

Because affiliates choose their own delivery, the table below records the documented chains rather than a single "Rhadamanthys method". The details are described at the level defenders need for controls, not as a recipe.

CampaignLure and first stageLater stages (defender view)Primary controls
TA547, Germany, 2024 (Proofpoint)Invoice emails impersonating a German retailer; password-protected ZIP containing an LNK fileLNK starts PowerShell, which fetches a remote script that decodes and loads the stealer in memory without writing it to diskBlock or quarantine LNK in archives, PowerShell logging and constrained language mode, ASR-style rules for script launch from archives
"CopyRh(ight)adamantys", 2024 (Check Point)Fake copyright-infringement notices from free webmail accounts, each target from a different sender, linking via a redirect to an archive on a file-sharing service; password in the emailArchive holds a legitimate executable, a malicious DLL and a decoy document; the DLL is side-loaded, padded to change its hash, persists through a registry Run key and injects into system processesInspect links to file-sharing archives, flag password-protected archives, detect DLL loads beside unexpected executables, alert on new Run keys
ClickFix campaigns, 2025 (Check Point)Fake verification or error pages asking the user to paste and run a commandNext stage retrieved in a noisy-looking PNG image, then staged modules downloaded from C2Restrict the Run dialog where not needed, script block logging, application control, targeted awareness. See the ClickFix and Fake CAPTCHA Paste-and-Run Attacks Guide

Two points recur across these chains. First, every chain needs the user to open, extract or paste something, so execution controls on the endpoint stop more infections than inbound filtering alone. Second, packers matter: MITRE ATT&CK's entry for the HeartCrypt packer-as-a-service (S9018) lists Rhadamanthys among the malware it has been used to pack, which is another reason sample hashes change constantly. For lure-specific depth, see the Malicious LNK Files Guide and the Malvertising and SEO Poisoning Guide.

Tactics and techniques (MITRE ATT&CK)

As of November 2026 ATT&CK v20 has no software entry for Rhadamanthys itself. The mapping below is built from the behaviours the cited vendor reports describe, using v20 technique IDs; treat it as an analyst mapping, not an official ATT&CK profile.

Behaviour reportedATT&CK techniqueWhere to see it
Attachment and link luresT1566.001 Spearphishing Attachment, T1566.002 Spearphishing LinkMail gateway, URL click logs
User opens LNK or archive contentT1204.002 Malicious FileProcess creation from archive extraction paths
Paste-and-run lureT1204.004 Malicious Copy and PasteExplorer spawning a shell or PowerShell, RunMRU registry entries
PowerShell loader running in memoryT1059.001 PowerShellScript block logging (Event ID 4104)
DLL side-loaded by a legitimate executableT1574.001 DLLImage load events for unsigned DLLs beside signed binaries
Hash-changing padding and packingT1027.001 Binary Padding, T1027.002 Software PackingLarge files with low entropy tails; packer verdicts
Payload hidden in image or audio filesT1027.003 SteganographyImage downloads followed by unexpected executable activity
Sandbox and analysis checksT1497.001 System ChecksSandbox detonation reports
Injection into Windows system processesT1055 Process InjectionSuspended process creation then remote memory writes
Persistence via registryT1547.001 Registry Run Keys / Startup FolderRegistry set events under Run keys
Browser and application credentialsT1555.003 Credentials from Web Browsers, T1552.001 Credentials In FilesNon-browser processes reading browser profile databases and wallet folders
WebSocket C2 and data uploadT1071.001 Web Protocols, T1041 Exfiltration Over C2 ChannelProxy and EDR network telemetry

Detection and hunting

Check Point's 0.9.x analysis recommends updating parsers and signatures because formats, mutex naming and configuration structure changed between versions, and focusing on behaviour. It also notes the malware queries several public time servers before contacting C2 and that configurations can list multiple C2 endpoints. Useful detections, in priority order:

  1. Credential store access by the wrong process: alert when a process other than the browser or password manager opens browser login and cookie databases, wallet directories or password manager vaults. This is the highest-value signal for any stealer family.
  2. Injection pattern: a process started suspended followed by memory writes into a common Windows binary, which Check Point names as a behaviour to alert on.
  3. Script delivery: PowerShell launched from an LNK inside a user's temporary archive extraction folder, or from Explorer after a Run dialog entry, with script block logging showing Base64 decoding and in-memory loading.
  4. Side-loading: a recently written, unsigned DLL loaded by a signed executable running from Downloads, Documents or a temporary folder, especially followed by a new Run key.
  5. Network pattern: a newly seen user-folder executable making several NTP or time-service lookups and then a WebSocket connection to a rare destination.
  6. Identity signals after the fact: new sign-ins from unfamiliar infrastructure using an existing session, password manager exports, or API keys from the device used from new networks. These catch infections the endpoint missed.

Proofpoint published Emerging Threats signature IDs for Rhadamanthys traffic (2854802, 2853002, 2853001 and 2043202) alongside its indicators; Check Point and Proofpoint both publish hashes and C2 addresses. Use them for retrospective hunting at least 90 days back, but expect them to age quickly. For takedown-related indicator handling, see the Cybercrime Law Enforcement Takedowns Guide.

MetricWhat good looks likeWhy it matters
Time from stealer detection to revocation of all sessions and tokens for the userHours, not daysStolen sessions are used quickly after harvest
Percentage of endpoints with script block logging and application control in enforce modeRising each quarter toward full coverageMost published chains rely on script or unsigned binary execution
Stealer-log exposure alerts closed with a full credential reset100 percentPartial resets leave tokens and API keys live
Developer machines whose local cloud and CI credentials are short-livedMajority, tracked by teamLimits what one infected laptop exposes
Users reporting copyright, invoice or verification lures before clickingReport rate rising, click rate fallingDelivery depends on user action

Prevention controls

  • Application control: enforce allowlisting so unsigned binaries in user-writable folders cannot run, and block LNK, script and HTA execution from archive paths. See the Application Allowlisting Guide.
  • PowerShell hardening: script block and module logging on, constrained language mode for standard users, and AMSI-integrated antimalware. See the PowerShell Security Guide.
  • Email and web controls: quarantine password-protected archives and LNK attachments, rewrite and sandbox links to file-sharing services, and treat unsolicited copyright or takedown notices as a high-risk theme.
  • Browser policy: disable the browser's built-in password saving on managed devices in favour of a managed password manager, and shorten session lifetimes for high-value SaaS.
  • Phishing-resistant and device-bound authentication: FIDO2 passkeys stop replay of stolen passwords; device-bound session controls, where your identity provider supports them, reduce the value of stolen cookies.
  • Developer secrets off disk: replace long-lived cloud keys and personal access tokens in local files with federated, short-lived credentials. See the Secrets Management Guide.
  • Wallet hygiene: keep recovery phrases offline, never photographed or scanned, given the OCR module's search for BIP-39 words.

Responding to a Rhadamanthys infection

  1. Isolate the host through EDR network containment and preserve a memory image if your process allows it; Check Point notes much of the payload lives in memory in custom formats, so disk images alone miss evidence.
  2. Establish the infection time from the first suspicious process, LNK or archive, and list every account used on the device since then, including personal browser profiles on corporate hardware.
  3. Revoke sessions first, then reset credentials: sign the user out of the identity provider and every high-value SaaS application, revoke refresh tokens, then reset passwords and re-register MFA from a clean device. See the Session Token Theft Response Playbook.
  4. Rotate non-human credentials found on the host: cloud access keys, API tokens, SSH keys, CI/CD and package registry tokens, VPN certificates. Check their audit logs for use from new locations since the infection time.
  5. Reimage the endpoint rather than cleaning it, because staged modules and injected processes make partial removal unreliable.
  6. Hunt laterally for the same lure, sender pattern, archive name or C2 indicators across mail, proxy and EDR telemetry.
  7. Monitor for the stolen data through your stealer-log monitoring service for months afterwards. Europol's Endgame page pointed victims to the Dutch police CheckYourHack service and Have I Been Pwned. See the Infostealer Logs Monitoring and Response Guide.

Operation Endgame and what it means for defenders

Europol coordinated the latest Endgame phase from its headquarters in The Hague between 10 and 14 November 2025, taking down 1,025 servers used by Rhadamanthys, the VenomRAT remote access trojan and the Elysium botnet. Europol said the infrastructure had infected hundreds of thousands of victims worldwide and held several million stolen credentials, and that many victims did not know their systems were infected. The main suspect for VenomRAT was arrested in Greece on 3 November 2025.

  • Do not assume the family is gone. A server takedown removes infrastructure, not the code or the customer base; keep behavioural detections and indicator hunts in place.
  • Use the victim services. The Endgame announcement directed victims to CheckYourHack and Have I Been Pwned; check both for your users and domains.
  • Expect affiliates to move. Customers of a disrupted stealer can rent another one. Behavioural detections for credential-store access cover the replacement as well. See the Lumma Stealer Guide and the StealC Infostealer Guide for comparable families.

Framework mapping

Control areaNIST CSF 2.0NIST SP 800-53 Rev 5ATT&CK mitigation or D3FEND
Block unauthorised executionPR.PS-05CM-7M1038 Execution Prevention, D3-EAL Executable Allowlisting
Malware detection on endpointsDE.CM-09SI-3, SI-4D3-PSA Process Spawn Analysis
Network monitoring for C2DE.CM-01SI-4M1042 Disable or Remove Feature or Program (unused script hosts)
Credential and session revocationRS.MI-02IA-5, AC-12Session termination and credential reset
User awareness of luresPR.AT-01AT-2M1017 User Training

Common failure patterns

  • Reimaging without revoking. The laptop is clean but the sessions, refresh tokens and API keys it held still work. The Snowflake breach showed how long stealer-harvested credentials stay usable when nobody rotates them.
  • Relying on hash blocklists. Padding, packing and per-build mutexes mean last month's indicators miss this month's samples.
  • Allowing password-protected archives through. The password in the email body defeats gateway scanning of the contents.
  • Ignoring personal browser profiles. Users sign into personal accounts on work devices, and the stealer collects both; the exposure of shared passwords crosses that line.
  • Leaving developer secrets in files. Long-lived keys in configuration files or environment files turn a single infection into a cloud or pipeline incident.

Practitioner checklist

  • Alert on non-browser processes reading browser credential and cookie stores, wallet folders and password manager vaults.
  • Enforce application control for user-writable paths and block LNK and script execution from archives.
  • Turn on PowerShell script block logging and forward Event ID 4104 to the SIEM.
  • Quarantine password-protected archives and inspect links to file-sharing archives.
  • Add copyright-notice, invoice and fake verification lures to awareness training and the phishing report flow.
  • Write a stealer playbook that revokes sessions and tokens before resetting passwords and covers non-human credentials.
  • Move developer cloud and CI credentials to short-lived, federated access.
  • Run retrospective hunts with published Rhadamanthys indicators and subscribe to stealer-log monitoring for your domains.

Standards and references

Related NHI Mgmt Group resources: Infostealer Threat Guide · Infostealers vs RATs · Session Hijacking and Cookie Theft Guide · Cybercrime Law Enforcement Takedowns Guide

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.

NHIMG Editorial Note
Researched and written by the NHI Mgmt Group editorial team using AI tools, under the editorial direction of Lalit Choda. Facts are drawn from the sources listed on this page and checked by our automated quality process. Last updated 14 November 2026. Spotted an error? Use our contact page.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM and Agentic AI security. nhimg.org