Tag: Dark Web

  • TSMC Leak Claim, Daxus Logs, Kazakhstan Court, Greek Ticketing, and Telegram Data

    TSMC Leak Claim, Daxus Logs, Kazakhstan Court, Greek Ticketing, and Telegram Data

    SOCRadar Dark Web Team identified several underground posts, including an alleged TSMC data leak, a Daxus credential marketplace advertising nearly 3 million records, and a claimed Kazakhstan Supreme Court database containing 507 million records. Other posts involved an alleged Greek citizen-service ticketing breach and a collection of 26.5 million Telegram user records.

    Daxus Advertises Alleged Credential Database

    daxus credential database sale

    SOCRadar Dark Web Team identified a dark web marketplace advertisement associated with Daxus, a vendor offering infostealer logs and credential-related services. The operation promotes a collection of approximately 2.96 million URL-login-password (ULP) records through its service platform, Daxus.pro.

    The advertised services include searchable credential logs, session cookies, private logs, and email account access. The listing does not establish which malware families collected the credentials or identify specific affected organizations. If authentic, these records could facilitate credential stuffing, account takeover, and unauthorized access to corporate services, particularly where passwords have been reused.

    Alleged Kazakhstan Supreme Court Database is Offered for Sale

    kazakhstan supreme court database sale

    SOCRadar Dark Web Team detected a dark web post advertising an alleged 507 million-record database belonging to Kazakhstan’s Supreme Court (Sud.kz). The seller, operating under the alias Brose, claimed the dataset contains information spanning 2014 to 2026 across 186 database tables, with an asking price of $50,000.

    The advertised contents include personal information on approximately 16.6 million citizens, 13.3 million case-party records, and 6.8 million company records. The actor also claimed possession of national identification numbers (IIN), passport details, court documents, password hashes, administrative credentials, session tokens, and payment information. If verified, the exposure could create substantial risks of identity theft, financial fraud, account compromise, and misuse of sensitive judicial records.

    Alleged TSMC Data Leak is Released

    tsmc data leak release

    SOCRadar Dark Web Team detected a dark web post claiming to release data allegedly belonging to Taiwan Semiconductor Manufacturing Company (TSMC). The threat actor, operating under the alias SmithBrianI, announced that the first batch of data was being made freely available after failing to attract a buyer.

    The post included a hidden download link and referenced a second batch scheduled for November 18, 2026, with a stated price of 5 BTC. The actor also mentioned AMD as a potential next target. However, no record count, data categories, or verifiable samples were disclosed. The alleged compromise remains unverified, and the extent of any potential exposure is unknown.

    Alleged Greek Ticketing Database Breach is Detected

    greek ticketing database breach

    SOCRadar Dark Web Team identified a post claiming unauthorized access to a database associated with Greece’s KEF citizen-service ticketing platform, referencing the domain superset-kef-test.ccticketing.eu. The actor, using the alias Noiem, claimed the dataset contains approximately 5.93 million records, involving around 1.27 million unique Greek tax identification numbers (AFM).

    The reportedly exposed information includes names, AFM identifiers, AMKA social security numbers, health insurance identifiers, email addresses, phone numbers, residential addresses, service requests, and password hashes. The listing also referenced beneficiary records, tax audit information, and more than 892,000 service tickets. If authentic, the data could enable identity fraud, targeted phishing, and impersonation of government services. The claims have not been independently verified.

    Alleged 26.5 Million Telegram User Records are Shared

    telegram 26m user records share

    SOCRadar Dark Web Team detected a dark web post advertising a collection of approximately 26.5 million Telegram user records covering the period between 2020 and 2024. The actor, operating under the alias Ghost_King, described the dataset as a compilation of five CSV files combining Telegram scraping data with records from Glaz Boga, FSBTool, and other sources.

    The sample information reportedly includes Telegram IDs, usernames, phone numbers, first and last names, and email addresses. The actor claimed to have removed duplicate entries and standardized phone numbers, focusing on users in Russia, Ukraine, Belarus, Moldova, Latvia, and neighboring countries. While the dataset could facilitate targeted phishing, account enumeration, and identity fraud, the listing does not establish a direct breach of Telegram’s infrastructure.

    Powered by DarkMirror™

    Gaining visibility into deep and dark web threats can be extremely useful from an actionable threat intelligence and digital risk protection perspective. However, monitoring all sources is simply not feasible, which can be time-consuming and challenging. One click-by-mistake can result in malware bot infection. To tackle these challenges, SOCRadar’s DarkMirror™ screen empowers your SOC team to follow up with the latest posts of threat actors and groups filtered by the targeted country or industry.

    Cti For Soc Analysts

    Source: SOCRadar · 2026-11-16

  • 354 dark web vendors and buyers arrested in Operation TorNado

    Europol and police in 14 countries arrested 354 dark web vendors and buyers in Operation TorNado, which ran from May 2025 to October 2026. Investigators went after people who had traded on five marketplaces that were already offline: Archetyp, Bohemia, Nemesis, Kingdom and Tor2Door. Seized money (Source: Europol) The people who used those markets were still exposed after the sites went dark. Police also went after vendors and buyers on forums, single-vendor shops and encrypted … More →

    Source: Help Net Security · 2026-11-13

  • Dark Web Profile: n0n Ransomware

    Dark Web Profile: n0n Ransomware

    n0n is a data-exfiltration and destructive-extortion group that surfaced on the Dark Web in September 2026 and scaled to 93 listed victims within roughly five weeks. Its operators exfiltrate data, encrypt or destroy infrastructure, and then work victims through keyed negotiation rooms, per-victim Bitcoin wallets and a public leak site branded with the slogan “n0n | no js | no mercy”.

    In September 2026, the SOCRadar Threat Research Unit (STRU) found an exposed staging server holding the live workspace of an n0n operator. The directory exposed the operation end to end: credential stuffing against FortiGate SSL-VPN appliances, hijacked firewall management centers, mass S3 deletion, a custom C++ locker, and an AI coding agent backed by more than $7,000 in paid LLM API usage. The findings tie the workspace to at least 23 organizations across 14 countries.

    This profile is based on the findings of our prior research into n0n Ransomware.

    Who Is n0n Ransomware?

    n0n has been actively monitored since September 2026. The group insists on the exact spelling of its name, “two lowercase n letters with a zero between them,” and its locker appends the .n0n extension to encrypted files.

    N0n Ransomware threat actor card
    N0n Ransomware threat actor card

    The operation grew quickly. An inspection of the leak portal on October 15, 2026, showed 93 targeted entities: 52 marked “exposed,” 21 “leaked,” 15 “pending,” and 5 “honored” after ransom payments. Tracking logged 93 victims across 27 countries, with the heaviest concentration in the United States and South Korea. As of October 19, total listed ransom demands reached $283.7M.

    n0n shows no ideological, hacktivist or state-directed motive. Every demand is financial, and target selection favors organizations whose compromise causes the most operational damage.

    How n0n Runs Its Extortion Operation

    n0n’s server codebase describes a three-stage extortion process, and its negotiation rooms present the same stages to victims:

    • Attack: Data is exfiltrated and systems are encrypted or destroyed while the incident stays private (“Nothing is public and only you know about the attack”)
    • Public exposure: The victim is listed on the leak site to raise pressure through media and search engine visibility
    • Files leaked: Archives are released through ZIP downloads and web-seeded torrents, the point at which the group reminds victims that “notification obligations typically apply”

    Leak Site and Negotiation Rooms

    The leak site runs as separate onion and clearnet builds, with keyed negotiation rooms for each victim. The clearnet domain n0n[.]to is now inactive; checks on October 15 and 19 used the onion address. Room paths were also used to serve payloads to compromised appliances.

    Public leak site of the n0n ransomware group
    Public leak site of the n0n ransomware group

    Victims must pay an upfront fee of 10% of the total ransom to access their negotiation portal, which prevents them from inspecting stolen files before paying. During the October wave, n0n dropped negotiation entirely: victims had to pay the listed address in full or face automatic publication when the countdown expired, 96 hours for mid-October listings and seven days for the maritime batch.

    Organizations that pay are rebranded as anonymous “HONORED – data destroyed” entries and removed from public rosters. Five such cards were logged by October 19, including one previously named US company and one apparent internal test entry, though no transactions have been validated on-chain.

    n0n negotiation room for FinSoft
    n0n negotiation room for FinSoft

    Pressure Tactics

    Extortion emails are sent direct to victims’ mail exchangers from notice@n0n-cases[.]net through custom Python scripts, bypassing standard mail providers. Automated countdown warnings reach union leadership, MSPs, clients and media outlets, with the line “Silence is read as refusal.” In the Transcom incident, a secondary alert went directly to PayPal’s security team, even though the compromised data sat only on Transcom’s network.

    Extortion email sent to PayPal for third-party leverage
    Extortion email sent to PayPal for third-party leverage

    Payment Infrastructure

    n0n is Bitcoin-only and generates a unique, single-use address for each victim. The operator derives these addresses with an Electrum wallet from a single BIP84 seed, using per-victim indexes. SOCRadar collected 55 addresses from ransom notes, maritime cases, operator derivation scripts and leak-site sources. A read-only check of all 55 found zero transactions, balances or pending mempool activity as of October 16, 2026.

    Leak Publication Pipeline

    Once a deadline passes, the operator flips the victim to “leaked.” Builder scripts compress data from the victim’s harvest folder, and every package passes two release gates first: guard.py blocks releases that contain malformed data or internal identifiers, and deepscan.py scrubs operator IPs, hostnames, credentials and AI agent strings. Archives are hosted on the leak site and seeded as torrents from a VPN-backed seedbox, and a public counter recorded 176,901 leak.zip downloads by October 19.

    n0n Campaign Timeline

    Date (2026) Activity
    September 17 Negotiation portals opened
    September 18 Network of Venezuelan ISP Inter severed
    September 23-25 First leak wave (10 organizations)
    September 25-26 Exposed staging server identified by STRU
    September 27-29 Second leak wave (3 organizations)
    October 1-2 Third leak wave (2 organizations)
    October 5-7 Two settlements marked “honored”
    October 12 38 maritime, port and logistics organizations listed
    October 19 All 38 maritime listings leaked
    October 22 Portals rebuilt
    n0n campaign timeline, September-October 2026
    n0n campaign timeline, September-October 2026

    Inside the Exposed n0n Operator Workspace

    STRU identified a Python HTTP file server at 185.157.161[.]148:8877 exposing the operator’s live working directory across two snapshots on September 25 and 26, 2026. Files were added, moved and deleted between snapshots, confirming active use:

    • 1,719 files captured across both snapshots
    • 829 scripts, including 589 Python, 28 PowerShell and 207 shell scripts
    • 5 C++ encryptor sources for Windows and Linux
    • An operational window of September 10-28, 2026, based on modification timestamps

    The operator’s infrastructure is segmented across four machines and supporting services:

    • Agent box (185.157.161[.]148): Central staging node hosting the AI coding agent and the kserver C2 panel behind Cloudflare
    • raas (188.214.135[.]130): Leak site backend, victim records, Electrum wallet, WinRM strike automation and C2 listeners
    • monster (81.161.239[.]10): Loot vault where all exfiltration terminates, leak builder, implant tools and an HTTPS C2 listener
    • Seedbox: Torrent seeding of all leak archives over a commercial WireGuard VPN, with a watchdog that resets stale tunnels

    Cover domains, Tor relays, an SSH pivot, residential proxies and third-party AI gateways insulate these core servers.

    n0n operator's infrastructure
    n0n operator’s infrastructure

    n0n Attack Lifecycle

    The operator applies the same lifecycle to every case, but the primary impact shifts between exfiltration, encryption, destruction and network disruption depending on the target.

    Initial Access Through Credentials

    STRU found no exploit payloads or CVE references in the workspace. Every documented intrusion relied on weak or compromised credentials:

    • FortiGate SSL-VPN credential stuffing: A 15-thread routine through residential proxy pools against an index of 305 appliances, with every request carrying the header Cookie: APBSID=n0n
    • RDP and WinRM: Sequential sweeps for default or weak passwords, driven by OCR-assisted RDP scripts
    • Cloud credentials: Long-lived AWS IAM keys with list, read and delete rights
    • Hosting provider sessions: Stored administrative browser sessions used to take over management, billing and monitoring portals
    • SaaS and CI/CD secrets: API tokens and cloud credentials harvested from source code and deployment pipelines
    FortiGate SSL-VPN credential checker used for credential stuffing
    FortiGate SSL-VPN credential checker used for credential stuffing

    Three cloud-database cases were staged under the label “gl-0day,” but no exploit code was found and their access vector remains unknown.

    Persistence and Credential Harvesting

    After access, the operator creates hidden administrator accounts disguised as services, such as svc_upd (“Software Update”) and svc_reg (“Registry Service”), with non-expiring passwords. Other persistence includes root resets through GRUB init=/bin/bash over OCR-driven VNC consoles, a gost SOCKS5 proxy disguised as a scheduled task, and OpenSSH on port 2222. Collection targets Chrome profiles and credential stores, session tokens, full IMAP mailboxes and incoming MFA codes.

    Creation of the hidden admin account svc_upd
    Creation of the hidden admin account svc_upd

    Exfiltration

    Each victim gets a dedicated pipeline with per-target byte budgets, from 1 GB to 110 GB per cloud bucket and an 8 GiB cap for in-band Windows exfiltration. Data leaves through Base64-encoded ZIP POSTs to C2, SigV4 requests over Tor, SSH-tunneled TAR archives and public file-hosting services, and always ends up on monster under /root/harvest/<victim>/.

    Encryption and Data Destruction

    The n0n locker runs inside wrapper scripts that stop database and security services, delete shadow copies and wipe free space. Destruction is a deliberate lever on its own: the operator purges source files after exfiltration, flips firewall rules to BLOCK, and in one case claimed to have removed 46 servers and their backups. The pitch to victims is direct: “the only restore path is the disk images we hold.”

    The n0n Encryptor

    The original C++ source of the locker was recovered from the staging server; the compiled binaries were not. The locker is a lightweight AES-256-CBC file encryptor with no network capability, key exchange, per-victim keys, ransom note logic, persistence or anti-analysis. All of that comes from external operator wrapper scripts.

    Variant Platform Used Against
    nenc.cpp (v1) Windows, CNG / AES-NI AFRICA-TECH; FinSoft
    nenc2.cpp (v2) Windows, CNG Staged against a medical-laboratory database server
    nenc2l.cpp / nenc2l_main.cpp Linux, OpenSSL EVP or tiny-AES-c Dediserve
    finenc3-5.ps1, denc.ps1, datasweep.ps1 PowerShell / .NET FinSoft
    FMC appliance stage Cisco FMC (Linux), OpenSSL Firewall-manager victims
    n0n execution chain on Windows, Linux and Cisco FMC
    n0n execution chain on Windows, Linux and Cisco FMC

    The Windows wrapper stages tooling in C:\Users\Public\n0n\, stops MSSQL, MySQL and W3SVC, disables Symantec Endpoint Protection, runs vssadmin delete shadows /all /quiet, wipes free space with cipher /w, and reruns encryption every 5 minutes as a SYSTEM scheduled task. It drops n0n_READ_ME.txt across roots, shares and desktops, sets a Run key that opens the note at logon, and finally overwrites the toolkit three times before deleting it.

    The Linux variant is the most cautious build: it writes to a .n0n.tmp file, decrypts and verifies the result against an MD5 checksum, and only then renames the file. At FinSoft, the first two PowerShell versions failed because a Base64 key decoded to 33 bytes, and a third version fixed the bug mid-attack.

    Each build uses a single hardcoded 32-byte key. nenc.cpp and the FinSoft PowerShell encryptor share an identical key, so a decryptor obtained by one victim would likely unlock the other’s files. Paid “decryptors” are this static key.

    AI-Assisted Operations

    The operator works with opencode, an open-source AI coding agent, configured with 154 model entries across 7 providers, including commercial gateways, “uncensored” and “abliterated” models with safety training removed, and a self-hosted Qwen 27B model on a rented GPU server. Billing records confirm active use of three gateways, and key-status checks show more than $7,216 in API spending as of October 22, 2026, with spending focused mainly on safety-removed models.

    The agent produced hundreds of numbered, iteratively patched scripts, stored encryption keys in its own memory file, and drove the C2 panel through saved session cookies. The configuration defines no automatic failover between models, and STRU observed a human-directed operation supported by AI, with no fully autonomous attack system.

    Decoy login page of the operator's kserver C2 panel
    Decoy login page of the operator’s kserver C2 panel

    Notable n0n Cases

    • Dediserve (cloud hosting, Ireland/Germany): Stolen admin browser credentials led to takeover of the provider’s management stack; the operator ran commands through Zabbix scripts and OnApp recipes named “disk usage check” and “maintenance check,” converted tenant backups into VM templates, and deployed the only Linux n0n locker variant
    • Inter (ISP, Venezuela): The operator ran inter_lock.py directly on the ISP’s Cisco Firepower Management Center, switched all access-control rules to BLOCK, and claimed to have severed national network traffic on September 18, 2026, alongside a 15 BTC demand
    n0n's note on its control of the Inter network
    n0n’s note on its control of the Inter network
    • Fanatics (sports e-commerce, US): An exposed AWS access key enabled exfiltration from 40 S3 buckets, followed by deletion of a bank-transaction archive and twelve confidential buckets without any encryptor; the group claimed 217.7 GB and demanded 30 BTC
    • ChiBiTek (managed IT services, US): Reached through the FortiGate credential-stuffing pipeline, profiled in operator scripts, and listed on the leak site on October 11

    n0n Targets and Victimology

    n0n scales initial access through RDP scanning and FortiGate credential stuffing, then selects targets from successful intrusions. High-profile entities were reached through deliberate scouting. Its typical target is a mid-sized organization with sensitive internal assets and edge services protected by weak or compromised credentials.

    The staging server’s data covers 23 organizations across 14 countries in 15 distinct sectors, with Software/SaaS and retail tied at three victims each. As of October 19, 15 of them were published on the leak site, 4 were compromised but not published, and for 4 more, access was attempted with no outcome shown. Demands ranged from 0.5 BTC for a government ministry to 33.04 BTC for the United Federation of Teachers.

    Geographic distribution of victims
    Geographic distribution of victims

    The operation extends well beyond that server. On October 12, 2026, n0n listed 38 maritime, port and logistics organizations across 18 countries in a single batch, with demands from $0.4M to $28M and a combined $275.8M. Maritime security research tied at least 21 of the 38 to customers of South Korean maritime software provider CyberLogitec, but CyberLogitec has not confirmed an incident, and one listing reuses another victim’s sample file, so the supply-chain link remains unverified. After the October 19 deadline, none of the 38 were marked settled, none of their payment addresses showed any transaction, and all were published.

    Sector distribution of victims
    Sector distribution of victims

    Attribution

    SOCRadar STRU assesses with medium confidence that n0n is a multi-operator, large-scale cybercrime operation, and that the exposed server is a private workspace belonging to core operators or administrators rather than hired affiliates. The evidence includes collective language in extortion notes, an operational tempo above single-operator bandwidth, the absence of the maritime campaign from this server’s artifacts, direct patch scripts for the leak site backend, and seed-level control of the wallet derivation tooling.

    No affiliate recruitment material or revenue-sharing logs were found, and the corpus offers no geographic or national attribution. All custom tooling is written in English; the only Cyrillic artifact is a localized desktop folder path on a victim endpoint.

    MITRE ATT&CK Techniques Used by n0n

    Tactic Technique Procedure
    Reconnaissance T1595.001 – Active Scanning: Scanning IP Blocks RDP batch scans and a ranked list of 305 FortiGate SSL-VPN appliances
    Initial Access T1078 – Valid Accounts Weak or default RDP/WinRM administrator credentials; firewall-manager admin accounts
    Initial Access T1078.004 – Valid Accounts: Cloud Accounts Exposed long-lived AWS IAM access keys
    Initial Access T1133 – External Remote Services Internet-facing FortiGate SSL-VPN, RDP and WinRM
    Execution T1059.001 – Command and Scripting Interpreter: PowerShell Strike, exfiltration and implant scripts pasted into RDP sessions
    Execution T1053.005 – Scheduled Task/Job: Scheduled Task Locker and audit runs as SYSTEM tasks every 5-10 minutes
    Execution T1651 – Cloud Administration Command AWS SSM commands, hosting-panel recipes and Zabbix script execution
    Persistence T1053.005 – Scheduled Task/Job: Scheduled Task AfricaSync, FinSoftSync*, SvcSync, NetAssist
    Persistence T1547.001 – Boot or Logon Autostart Execution: Registry Run Keys n0nNote opens the ransom note at every logon
    Persistence T1136.001 – Create Account: Local Account Hidden administrators svc_upd, svc_reg, svc<slug>
    Persistence T1098.004 – Account Manipulation: SSH Authorized Keys Operator keys installed via VNC console on cloud clones; root@3090-AI on FMC appliances
    Defense Evasion T1562.001 – Impair Defenses: Disable or Modify Tools Symantec Endpoint Protection stopped and disabled
    Defense Evasion T1564.002 – Hide Artifacts: Hidden Users SpecialAccounts\UserList value 0 for the cover accounts
    Defense Evasion T1070.004 – Indicator Removal: File Deletion Toolkit overwritten three times and deleted; launchers self-delete
    Credential Access T1110.004 – Brute Force: Credential Stuffing FortiGate /remote/logincheck credential stuffing via residential proxies (APBSID=n0n)
    Credential Access T1110.002 – Brute Force: Password Cracking Offline cracking of stolen SaaS user hashes and encrypted 7z archives
    Credential Access T1555.003 – Credentials from Password Stores: Credentials from Web Browsers Chrome/Edge stores decrypted, including Chrome’s app-bound key
    Credential Access T1539 – Steal Web Session Cookie Live administrator Chrome profiles copied and driven over DevTools
    Credential Access T1111 – Multi-Factor Authentication Interception Billing-panel MFA codes read from the staff mailbox
    Credential Access T1003.003 – OS Credential Dumping: NTDS NTDS.dit copy from a shadow copy (scripted)
    Discovery T1046 – Network Service Discovery Internal sweeps of 445/3389/5985/1433
    Discovery T1580 – Cloud Infrastructure Discovery EC2/ECS/SSM mapping and hosting-panel VM inventories
    Discovery T1619 – Cloud Storage Object Discovery S3 bucket listing and prefix census
    Lateral Movement T1021.001 – Remote Services: Remote Desktop Protocol Headless xfreerdp sessions driven by key injection and OCR
    Lateral Movement T1021.006 – Remote Services: Windows Remote Management pywinrm strikes, triage and browser theft
    Collection T1530 – Data from Cloud Storage S3 bulk pulls with per-bucket byte caps
    Collection T1114.002 – Email Collection: Remote Email Collection Full IMAP dump of a provider staff mailbox
    Collection T1115 – Clipboard Data RDP results returned through the clipboard
    Collection T1113 – Screen Capture Remote screen captured and read with OCR to drive logons
    Command and Control T1071.001 – Application Layer Protocol: Web Protocols HTTPS implant C2 (c2srv.py) and the kserver panel
    Command and Control T1001 – Data Obfuscation Status beacon disguised as card-payment JSON to /x/i
    Command and Control T1572 – Protocol Tunneling SSH -D SOCKS tunnels over Tor
    Command and Control T1090.003 – Proxy: Multi-hop Proxy Tor, SSH pivots, gost exits on compromised hosts, residential proxies
    Exfiltration T1041 – Exfiltration Over C2 Channel Zip-base64 POSTs to /c2/…/out (8 GiB budget)
    Exfiltration T1567.002 – Exfiltration Over Web Service: Exfiltration to Cloud Storage Uploads to a public file-hosting service
    Impact T1486 – Data Encrypted for Impact n0n locker (Windows/Linux) and the OpenSSL stage on FMC appliances
    Impact T1485 – Data Destruction S3 mass deletion; a provider’s servers and backups removed
    Impact T1490 – Inhibit System Recovery Shadow copies deleted; free space wiped
    Impact T1489 – Service Stop MSSQL, MySQL and IIS stopped before encryption
    Impact T1565.001 – Data Manipulation: Stored Data Manipulation Firewall rules flipped to BLOCK in the FMC configuration database

    n0n Indicators of Compromise (IOCs)

    The full IOC set is available in the SOCRadar STRU threat intelligence repository.

    How to Defend Against n0n Ransomware

    • Harden remote access: Enforce phishing-resistant MFA on RDP, SSL-VPN and admin portals, monitor /remote/logincheck for credential stuffing, and invalidate credentials found in infostealer logs
    • Lock down firewall management: Restrict FortiManager and FMC interfaces with IP allow-lists and MFA, require formal approval for policy changes, remove shared or default credentials, and flag REST token creation from unknown IPs
    • Reduce cloud key risk: Replace long-lived cloud keys with short-lived roles, keep list, read and delete S3 permissions off a single IAM identity, enable Object Lock and versioning, and alert on high-volume DeleteObjects calls
    • Keep offline, immutable backups: n0n purges shadow copies, zero-fills free space and deletes provider-hosted backups, so air-gapped or immutable copies are the reliable recovery path
    • Isolate administrative browsers: Hosting and SaaS operators should run management portals from isolated workstations with hardware-backed session controls, and keep MFA passcodes out of local admin mailboxes
    • Hunt for n0n behavior: Search EDR, CloudTrail, WinRM/RDP and firewall logs for the IOCs above and for C:\Users\Public\n0n\, .n0n renames after MSSQL/W3SVC/MySQL stops, SYSTEM tasks such as SvcSync or NetAssist, Chrome launched with –remote-debugging-port, Snort rule “n0n-lock” (sid 2100001), authorized_keys ending in root@3090-AI, and JSON POSTs to /x/i containing “ippay-web”

    Frequently Asked Questions

    What Is n0n Ransomware?

    n0n is a financially motivated data-exfiltration and destructive-extortion group active since September 2026. It encrypts files with its own AES-256-CBC locker, appends the .n0n extension, destroys data and backups, and publishes stolen files on a Tor and clearnet leak site. By October 19, 2026, it had listed 93 victims with $283.7M in combined ransom demands.

    How Does n0n Gain Initial Access?

    Every intrusion documented in the operator’s workspace used weak or compromised credentials. The main vectors were FortiGate SSL-VPN credential stuffing against 305 appliances, RDP and WinRM password sweeps, exposed AWS IAM keys, hijacked hosting-provider browser sessions and secrets taken from SaaS and CI/CD pipelines. No exploit code or CVE references were found.

    Does n0n Use AI in Its Attacks?

    Yes. The operator uses the opencode AI coding agent with 154 configured models, including uncensored and safety-removed models and a self-hosted Qwen 27B model, and spent more than $7,216 on LLM APIs by October 22, 2026. The agent wrote and patched hundreds of attack scripts, but the operation remains human-directed.

    Can Files Encrypted by n0n Be Recovered?

    Each n0n build uses a single hardcoded 32-byte key, and the same key appears in two builds used against different victims, so one victim’s decryptor would likely decrypt another’s files. The wrapper scripts still delete shadow copies, wipe free space and overwrite originals, which makes offline or immutable backups the reliable recovery option.

    Who Does n0n Target?

    n0n targets organizations ranging from small IT firms to global enterprises across at least 27 countries, with the highest concentration in the United States and South Korea. Victims include SaaS vendors, retailers, ISPs, cloud providers, a teachers’ union, a government ministry, pharmaceutical and gambling companies, and 38 maritime, port and logistics organizations listed in a single October batch.

    Is n0n Linked to a Specific Country?

    No. SOCRadar STRU found no geographic or national attribution in the operator’s files. All custom tooling is written in English, and the only Cyrillic artifact was a localized folder path on a victim’s endpoint.

    Threat Actor

    Source: SOCRadar · 2026-11-12

  • Dark Web Profile: CyberXero

    Dark Web Profile: CyberXero

    CyberXero is a Russian-speaking, financially motivated Initial Access Broker (IAB) active on Dark Web forums, including Dread and HackForums. The actor runs two parallel operations: an opportunistic global campaign that mass-exploits WordPress, Magento and other e-commerce platforms for access and payment data, and a directed campaign that conducts curated reconnaissance against Ukrainian energy and utilities organizations.

    What sets CyberXero apart is its AI layer, with up to 51 specialized Claude Code agents on its primary workstation and a PentAGI deployment wired into a Cobalt Strike Team Server through the Anthropic API. The SOCRadar Threat Research Unit has tracked the actor’s infrastructure since July 2026, and CyberXero remained active at the time of publication.

    This profile is based on the findings of our prior research into CyberXero.

    Who Is CyberXero?

    CyberXero came to light after SOCRadar identified an exposed open directory on an attacker-controlled HTTP server. The server held more than 90,000 files and 3,000 subdirectories from an active Linux workstation, including engagement folders named after target domains, exploitation toolkits, AI agent configurations, session logs and victim database exports.

    CyberXero threat actor card
    CyberXero threat actor card

    The actor’s own Dread post describes a progression from having no prior knowledge to dumping emails and password hashes from vulnerable databases, running into tokenization while trying to obtain credit card data, and moving toward digital skimming. That account matches the order of the compromises documented in its infrastructure, although the actor’s five to nine years of web hosting administration contradicts the claim of having “started from scratch”.

    CyberXero’s activity centers on three revenue streams:

    • Access brokering: Compromised WordPress administration panels, webshells and network footholds, consistent with an IAB model built around selling access
    • Payment and personal data theft: Database dumps, payment tables and a card skimmer endpoint tested against a production payment flow
    • Cryptocurrency payments: A self-hosted payment gateway that accepts Bitcoin, Litecoin, Ethereum, USDT (ERC-20 and TRC-20) and Monero

    From an Open Directory to the CyberXero Alias

    Attribution to the CyberXero alias rests on three independent lines of evidence. The open directory contained the operator’s Claude Code account data, the session logs showed the actor citing a Dread post written by the CyberXero account as proof of “owning” the forum, and external verification confirmed that the post was published by CyberXero and describes the same TTP progression observed in the infrastructure.

    Five independent identity anchors were collected across Anthropic, Apple, Google, Twitter/X and OpenAI accounts, and a record on a Russian freelance marketplace that requires identity verification adds further weight. Two additional Telegram accounts receive notifications from the actor’s operational bot and are assessed as either a second alias or a trusted collaborator; their exact relationship to the primary operator could not be resolved with the available data.

    The actor works in the Moscow time zone (UTC+3), supported by a provisioning timezone directive and 1,269 timestamped activity events peaking at 22:00 MSK, with virtually no activity between 02:00 and 08:00. Russian is the primary operational language, confirmed through script comments, a custom Claude Code instruction file and a mid-session language switch after a model refusal.

    What Are the Targets of CyberXero?

    SOCRadar identified more than 40 affected organizations across Ukraine, Poland, China, Pakistan and the Netherlands, excluding the 4,500+ WordPress sites hit through mass exploitation.

    Geographic distribution of confirmed CyberXero victims
    Geographic distribution of confirmed CyberXero victims

    Ukraine carries the highest concentration, with 14 or more entities identified at different stages of the compromise chain, mostly in the energy and utilities sector, plus four organizations in telecommunications, e-commerce and CRM compromised through the opportunistic pipeline. Seven Ukrainian energy and utilities entities, including the national transmission system operator and the country’s largest private energy holding, were subjected to curated reconnaissance; one file alone lists 95 enumerated subdomains of the two largest, covering mail, VPN, network dispatch and data platforms.

    Poland follows with four to five affected entities, two of them confirmed e-commerce compromises, and 284 scanned Polish domains with a heavy concentration of government institutions. The actor also compromised a Chinese corporate network cluster spanning at least three internal segments, attempted WAF bypass and credential spraying against two Pakistani national security government entities without confirmed success, and accessed a Dutch company through an exposed Magento configuration.

    Sector distribution of confirmed CyberXero victims
    Sector distribution of confirmed CyberXero victims

    Energy and utilities account for 28% of confirmed victims, followed by e-commerce and payment (24%), government (20%), ISP and telecom (16%), corporate networks (8%) and CRM and SaaS (4%).

    The scale of the confirmed impact stands out in two figures:

    • More than 628,000 records: Personal data of Ukrainian individuals in the actor’s possession, including home addresses and login IP addresses of residents of Kharkiv, a city on an active war front
    • 429 administration panels: WordPress panels confirmed accessible in a single automated run that scanned 4,708 targets and deployed shells on 32 hosts in 61 seconds

    The opportunistic pipeline fits a conventional IAB model, while the focus on Ukrainian critical infrastructure has no obvious direct monetization path. The most plausible explanation is access brokering to a buyer with a specific interest in that infrastructure, although no evidence of a sale, buyer communication or tasking was found.

    What Are the Techniques of CyberXero?

    CyberXero runs an opportunistic pipeline and a targeted pipeline from the same workstations, sharing part of the same arsenal during the same period. Its infrastructure spans eight correlated nodes across three ASNs, linked through unique provisioning tokens, a shared internal hostname, command histories and staging logs.

    Initial Access Through Mass WordPress Exploitation

    The actor’s in-house tool, wp2shell, is a five-module Python package that abuses a desynchronization in the WordPress REST API batch endpoint (/wp-json/batch/v1), where the validation and execution handlers apply different authorization contexts. It adaptively selects union-based, error-based or time-based blind SQL injection, writes a new administrator directly to the database, logs in and deploys a WSO-family webshell registered as an active plugin.

    Rogue administrators follow the pattern wp2_ plus eight random hexadecimal characters. A companion script, sc.py, works through the target queue and pushes each compromise to a rotating Telegram bot with the panel URL, credentials and shell status. Across the actor’s files, more than 56,000 URLs were scanned, 500 hosts were exploitable and 72 rogue administrator accounts were created during the primary observation period.

    Targeted Intrusions Against Ukrainian Organizations

    The targeted pipeline relies on deeper, manual chains. Against a Ukrainian classifieds and e-commerce portal, the actor combined blind SQL injection to recover an administrator email, a time-constrained password reset token extraction and a PHP deserialization chain built with phpggc, reaching remote code execution and dumping MySQL, MongoDB and Redis databases with 24,132 active sessions and 864 API tokens.

    Against Kharkiv’s district heating service, no exploitation was needed: a hardcoded administrator credential let a multithreaded scraper pull 564,073 subscriber records and 213,340 login log entries. Other confirmed exfiltration hit a regional ISP (63,621 customer records and 75,811 cleartext access credentials) and a gas provider (27 employee accounts with active session tokens).

    AI Augmentation and Safety Evasion

    CyberXero uses AI in two independent layers. On its primary workstation, a swarm of 51 Claude Code agents covers offensive, reconnaissance and post-exploitation roles, from c2-operator and exploit-chainer to scada-attacker and data-exfiltrator. On the Team Server, PentAGI connects to Cobalt Strike through the Anthropic API, with the payload-generation role assigned the highest token limit and reasoning level.

    A dedicated scope-guard agent wraps every task in a fictional “authorized pentest on own infrastructure” framing to lower the chance of model refusals, and the same ownership claim appears verbatim across four unrelated victims. When a model still refused, the actor reopened the task in a fresh session with the framing preloaded. The logs record repeated refusals nonetheless, including four declined requests to scan Dread, where the model asked for token-based ownership verification and rejected the actor’s attempt to paste the token back into the chat.

    Rapid Weaponization of Vulnerabilities

    CyberXero turns disclosed vulnerabilities into working exploits quickly. CVE-2026-4815 in Support Board 3.8.7 was published in June 2026, and a functional exploit with proxy rotation and automatic timing calibration appears in files dated July 2026, a cycle of less than 30 days. The arsenal also includes CVE-2026-63030 and CVE-2026-60137 in WordPress and CVE-2015-1397 in Magento.

    Command and Control With Cracked Cobalt Strike

    The actor runs Cobalt Strike 4.9.1 from a pirated build distributed through a Telegram channel, with a “Major Cobalt Strike” TLS certificate shared by about 184 other hosts that should not be treated as CyberXero infrastructure. The C2 receives HTTPS beacons on port 1002, serves HTTP staging on port 8044 and delivers PowerShell stagers on port 8033.

    Lateral Movement in a Chinese Network Cluster

    A separate campaign gained access to a Chinese corporate network by abusing a misconfigured Redis server through a rogue master-replication technique that planted the actor’s SSH key. From there, beacons enumerated the network, brute-forced a ZTE gateway with default credentials, moved laterally over WMI through a passwordless account, checked hosts for EternalBlue patch status and created the UpdSvc Windows service for persistence. Successful EternalBlue exploitation was not confirmed in the available logs.

    Monetization and Competitor Suppression

    A cryptocurrency payment gateway built on FastAPI and Celery, with local Litecoin and Monero full nodes, receives payments for sold access and data. The actor also built a card skimmer endpoint against a production payment flow, which had reached functional testing with about 50 masked card records. A script named abuse_monitor.py files abuse reports against competitors to URLhaus, VirusTotal and hosting providers under a fake fraud-analyst persona, a TTP undocumented in comparable IAB profiles.

    CyberXero’s opportunistic and directed pipelines, orchestrated through the AI layer
    CyberXero’s opportunistic and directed pipelines, orchestrated through the AI layer

    How Can Organizations Defend Against CyberXero?

    The controls below map to CyberXero’s documented TTPs, ordered along its intrusion chain from initial access to post-compromise activity.

    Hunt for Rogue WordPress Administrators

    Review WordPress user tables for any account whose username matches wp2_ followed by eight hexadecimal characters, and check the plugins directory for paths matching wp2shell_ with the same suffix. Either artifact indicates that wp2shell touched the host, even if the webshell is no longer active, so treat the site as compromised and rotate all administrator credentials.

    Patch Exposed Web Applications Within Days

    With a weaponization cycle under 30 days, patch windows measured in months leave WordPress, Magento, PrestaShop and Support Board deployments exposed. Prioritize CVE-2026-63030, CVE-2026-60137 and CVE-2026-4815, and confirm that configuration files such as Magento’s app/etc/env.php are not publicly reachable.

    Remove Hardcoded and Default Credentials

    Several CyberXero intrusions needed no exploit at all. Audit administration panels for shared or hardcoded credentials, replace default passwords on network devices such as gateways and routers, and eliminate local accounts with blank passwords that enable WMI-based lateral movement.

    Lock Down Internet-Facing Redis

    Do not expose Redis to the internet, require authentication, and disable or rename the SLAVEOF/REPLICAOF command where replication is not needed. Monitor authorized_keys files on Redis hosts for unexpected changes.

    Block and Hunt Known C2 Infrastructure

    Block the IP addresses listed in the IOC section and search historical logs for outbound connections to 42.193.227.214 on ports 1002, 8044 or 8033. Hunt for the UpdSvc service and PowerShell download cradles on Windows endpoints.

    Treat AI Agent Session Logs as Sensitive Data

    The CyberXero investigation shows that AI agent session logs record every planned task, target and workaround, making them the equivalent of a keylogger of the operator’s activity. Organizations running AI agents with terminal, network or data access should apply the same retention, encryption and access controls to these logs as to credential stores, and audit them for repeated attempts at the same refused task.

    Integrate Dark Web and Threat Intelligence Monitoring

    CyberXero advertises on Dread and HackForums, and IAB listings often surface before access is used. Monitoring Dark Web forums and Telegram channels for references to your domains, brands and infrastructure gives an early signal that access may be for sale. SOCRadar Threat Intelligence’s Advanced Dark Web Monitoring (ADWM) module provides this coverage.

    What Are the MITRE ATT&CK TTPs of CyberXero?

    Tactic Technique ID
    Reconnaissance Active Scanning: Vulnerability Scanning T1595.002
    Reconnaissance Gather Victim Network Information: DNS T1590.002
    Reconnaissance Gather Victim Org Information T1591
    Resource Development Acquire Infrastructure: Virtual Private Server T1583.003
    Resource Development Obtain Capabilities: Tool T1588.002
    Resource Development Obtain Capabilities: Exploits T1588.005
    Initial Access Exploit Public-Facing Application T1190
    Initial Access Valid Accounts: Local Accounts T1078.003
    Initial Access Brute Force: Password Spraying T1110.003
    Execution Command and Scripting Interpreter: PowerShell T1059.001
    Execution Command and Scripting Interpreter: Unix Shell T1059.004
    Execution Windows Management Instrumentation T1047
    Execution Software Deployment Tools T1072
    Persistence Server Software Component: Web Shell T1505.003
    Persistence Create or Modify System Process: Windows Service T1543.003
    Persistence Account Manipulation: SSH Authorized Keys T1098.004
    Persistence Create Account: Local Account T1136.001
    Stealth Obfuscated Files or Information T1027
    Stealth Masquerading: Match Legitimate Resource Name or Location T1036.005
    Credential Access OS Credential Dumping: LSASS Memory T1003.001
    Credential Access OS Credential Dumping: Security Account Manager T1003.002
    Credential Access Input Capture: Web Portal Capture T1056.003
    Credential Access Steal Web Session Cookie T1539
    Discovery Network Service Discovery T1046
    Discovery System Information Discovery T1082
    Discovery System Network Configuration Discovery T1016
    Lateral Movement Use Alternate Authentication Material: Pass the Hash T1550.002
    Lateral Movement Remote Services: SMB/Windows Admin Shares T1021.002
    Collection Data from Local System T1005
    Collection Data from Information Repositories T1213
    Collection Data from Cloud Storage T1530
    Command and Control Application Layer Protocol: Web Protocols T1071.001
    Command and Control Web Service: Bidirectional Communication T1102.002
    Command and Control Proxy: Multi-hop Proxy T1090.003
    Exfiltration Exfiltration Over C2 Channel T1041
    Exfiltration Exfiltration Over Web Service: Exfiltration to Cloud Storage T1567.002
    Impact Data Manipulation: Stored Data Manipulation T1565.001

    What Are the IOCs of CyberXero?

    Type Indicator Description
    IP 46.21.250.135 Primary workstation, open directory seed node
    IP 45.88.106.2 Provisioning server (port 1500)
    IP 212.193.31.189 Cryptocurrency payment gateway
    IP 42.193.227.214 Cobalt Strike Team Server and PentAGI
    IP 91.208.184.148 Secondary workstation
    IP 45.88.106.78 Mass scanning server
    IP 49.234.12.182 Staging and payload delivery server (Chinese cluster)
    IP 42.193.100.94 Redis exploitation node and PentAGI agents
    IP:Port 42.193.227.214:1002 Cobalt Strike HTTPS beacon
    IP:Port 42.193.227.214:8044 Cobalt Strike HTTP staging
    IP:Port 42.193.227.214:8033 Cobalt Strike PowerShell stager delivery
    SHA256 92a789444708fa1cb4cc5a89e0aa6cc7a279b62a7b8a4d2857255a18197d0090 stager32.bin, Cobalt Strike x86 stager (4.9.1 pirated build)
    MD5 3870a559b5b6ea6076f5f90029c180b7 stager32.bin, Cobalt Strike x86 stager
    SHA256 a5ae0aab352871bc0b038b3aa43b03eb223425628c884af7b9fc592cd34eb86c Obfuscated WSO webshell
    SHA256 71e21094c1ac1cf0275c91ed377965e248bca1c12711f6dd20e2682681fc1192 1.bin, Chinese cluster staging payload
    SHA256 59c535f47ab4d35f6d9fc8b8aec4a72438ea0b89e5a4dcea74b05d2d32cfa483 config.bin, Chinese cluster staging payload

    Behavioral indicators, weaponized CVEs and YARA rules are covered in the full threat research report. For additional IOCs and detection rules, you can visit the CyberXero threat actor profile on the SOCRadar platform.

    Threat Actor

    Source: SOCRadar · 2026-11-10

  • Canadian Webmail Access, US Bank Dump, Peru Leak, Spencer’s Breach, and ZoomInfo Sale

    Canadian Webmail Access, US Bank Dump, Peru Leak, Spencer’s Breach, and ZoomInfo Sale

    SOCRadar Dark Web Team identified several underground posts, including an alleged webmail access sale targeting a Canadian infrastructure company, a claimed US banking database dump involving sensitive credentials, and an alleged Peruvian government registry leak affecting more than 20 million records. Other posts advertised an alleged Spencer’s retail database breach and millions of business contact records reportedly sourced from ZoomInfo.

    Alleged Canadian Infrastructure Webmail Access Is Offered for Sale

    canadian infrastructure webmail access sale

    SOCRadar Dark Web Team detected a dark web post advertising alleged webmail access to a Canadian infrastructure company. The threat actor, operating under the alias Horsefucker, claimed that the targeted organization generates more than $10 million in annual revenue.

    The listing identified Canada as the target country and webmail as the access type but did not disclose the organization’s name, email domain, or supporting evidence. If legitimate, compromised webmail access could facilitate Business Email Compromise (BEC), credential theft, phishing, and further unauthorized access to corporate systems. The claim remains unverified.

    Alleged US Bank Database Dump and Cloud Credentials Are Advertised

    us bank database dump cloud credentials

    SOCRadar Dark Web Team identified a post claiming to offer a comprehensive database dump from an unnamed US financial institution. The actor, known as huanhoahong, advertised production and staging databases containing customer information, loan records, transaction histories, and financial data.

    The listing also claimed access to 492 AWS Secrets Manager values, approximately 7,000 sensitive keys, Terraform state files, Jenkins credentials, SSH private keys, and AWS administrative privileges. Sample records reportedly included names, Social Security numbers, dates of birth, and contact details. If authentic, the exposure could enable financial fraud, identity theft, and further compromise of the institution’s cloud infrastructure. The claims have not been independently verified.

    Alleged Peru Government Registry Leak Exposes 20M+ Records

    peru government registry leak 20m records

    SOCRadar Dark Web Team detected a dark web post in which a group identifying itself as MortemTeam claimed to have leaked more than 20 million records associated with the Peruvian government’s national registry systems.

    The advertised database reportedly contains DNI national identity numbers, RUC taxpayer identifiers, and physical addresses. The group presented the release as an attack against Peruvian government systems, although no specific compromised agency or intrusion method was identified. If authentic, the exposed information could support identity fraud, impersonation, and targeted phishing. The alleged leak remains unconfirmed, with no publicly accessible sample provided in the observed post.

    Alleged Spencer’s Database Breach Is Detected

    spencers database breach 461m records

    SOCRadar Dark Web Team identified a post claiming that US retailer Spencer’s had suffered a database breach. The threat actor, operating under the alias scarlettgroup, claimed to have released approximately 461.8 million structured database records, including information associated with more than 1.55 million customer accounts.

    The advertised dataset reportedly includes customer names, email addresses, phone numbers, dates of birth, password hashes, loyalty card information, transaction histories, payment records, and marketing data. If verified, the exposure could facilitate credential stuffing, loyalty fraud, phishing, and identity theft. The reported total represents database records across multiple categories rather than 461.8 million individual customers. The breach claim remains unverified.

    Alleged ZoomInfo Business Lead Database Is Offered for Sale

    zoominfo business lead database sale

    SOCRadar Dark Web Team detected an underground listing advertising approximately 30 million business leads allegedly associated with ZoomInfo, a business intelligence and contact data platform. The seller, operating under the alias LeadsGig, promoted the dataset as fresh US business contact information and provided a sample link.

    The advertised fields include names, business email addresses, company phone numbers, job titles, seniority levels, company locations, industry classifications, revenue information, and employee counts. If authentic, the dataset could facilitate targeted phishing, business email compromise (BEC), and corporate impersonation. However, the listing does not establish whether the information originated from unauthorized access, scraping, or previously distributed datasets.

    Powered by DarkMirror™

    Gaining visibility into deep and dark web threats can be extremely useful from an actionable threat intelligence and digital risk protection perspective. However, monitoring all sources is simply not feasible, which can be time-consuming and challenging. One click-by-mistake can result in malware bot infection. To tackle these challenges, SOCRadar’s DarkMirror™ screen empowers your SOC team to follow up with the latest posts of threat actors and groups filtered by the targeted country or industry.

    Cti For Soc Analysts

    Source: SOCRadar · 2026-11-09

  • Qilin Ransomware Admin Haise Allegedly Unmasked

    Qilin Ransomware Admin Haise Allegedly Unmasked

    GangExposed, the anonymous leaker who spent 2025 exposing Conti and Trickbot leadership, has moved on to Qilin. A new post on the channel shares an investigation naming a Russian national as “Haise”, the persona that has fronted Qilin’s affiliate program on underground forums since 2022.

    Only weeks ago, Germany arrested a Russian national suspected of being a leading Qilin member after extradition from Japan. Nothing public connects that suspect to Haise, and no authority has confirmed GangExposed’s identification.

    What we have is a single investigator’s chain of forum records, Telegram history and leaked Russian databases, and it deserves a careful read before anyone treats the name as settled.

    Photos of the individual GangExposed names as Haise, as published in the investigation; blurred and redacted by SOCRadar
    Photos of the individual GangExposed names as Haise, as published in the investigation; blurred and redacted by SOCRadar
    Key Takeaways
    • GangExposed published an investigation naming a Russian national as Haise, Qilin’s long-standing administrator persona
    • The identification rests on forum profiles, Telegram history and leaked Russian databases, and has not been independently verified
    • Haise’s role in Qilin is well documented, dating back to the persona’s arrival on RAMP in May 2022
    • No public source links Haise to the suspected leading Qilin member arrested in Germany in October 2026
    • Qilin keeps operating, with more than 2,350 known victims across 62 countries

    Haise Before the Leak

    Haise is one of the better-documented personas in the RaaS ecosystem. The profile joined the Russian-language RAMP forum on May 29, 2022, shortly before the group surfaced as Agenda, and was openly advertising Qilin’s affiliate program by February 13, 2023.

    Over the following years, Haise has been the voice of the program. In March 2023, researchers who infiltrated Qilin reached the persona over Tox and learned that affiliates kept 80% of ransom payments up to $3 million and 85% above that.

    In July 2023, Haise explained on RAMP that ransoms land in affiliates’ wallets first, with the operators’ share forwarded afterward. In June 2025 came the “call lawyer” feature, pitched to affiliates as legal support for squeezing victims during negotiations. A month later, a Telegram channel impersonating Europol offered rewards for Haise and a second administrator persona, “XORacle”. Europol called it a scam.

    So the public role is not in question. The identity behind it is what GangExposed claims to have solved.

    How the Identification Was Built

    The write-up describes “a chain of data linking the alias Haise to a specific individual” that “runs in one direction and is corroborated in reverse”, and closes with “a high degree of confidence” in the result.

    We are not reproducing the personal data in the investigation, but its logic can be summarized. It starts with RAMP approval logs from June 2022 that list “haise” among new members, and matches the persona to profiles on Exploit and XSS.

    A shared second alias across those profiles, plus a Telegram handle and an email address listed on them, carries the trail to other forums, including a dump of roughly 25,000 accounts from a third forum. From there the investigation moves into leaked Russian data: the email appears in a public services database that points to a named individual, and a relative’s leaked records lead to a phone number tied to a Telegram account, cross-checked against a leaked mobile subscriber database.

    That Telegram account is where the story gets more texture. It was active in public groups from September 15, 2018, to July 29, 2024, and has since been deleted. The messages read like someone running ransomware operations rather than dabbling: recruiting penetration testers for Windows, Linux, ARM, NAS and ESXi in April 2023, asking for forums with pentesters “apart from Exploit, XSS, and RAMP”, digging into how the ESXi kernel works, asking how to bypass and disable a major EDR product in August 2023, and asking “Who works with corporate targets?”

    Qilin advertisements and surrounding Telegram group messages cited in the investigation
    Qilin advertisements and surrounding Telegram group messages cited in the investigation

    The write-up also points to Qilin ads posted in the same groups on July 22, July 29 and August 7, 2024, and says the Qilin brand appears nowhere else in those groups. On July 29, 2024, the account replied within seconds to a group post about reporting that Scattered Spider had added Qilin and RansomHub to its toolkit.

    How the Chain Runs
    1
    RAMP approval logs June 2022
    “haise” listed among newly approved members
    2
    Exploit and XSS profiles
    Matched to the persona and sharing a second alias
    3
    Forum contact details
    A Telegram handle and an email address lead to other forums, including a dump of roughly 25,000 accounts
    4
    Leaked public services database
    The email points to a named individual
    5
    Relative’s leaked records
    A phone number tied to a Telegram account, cross-checked against a leaked mobile subscriber database
    6
    Telegram activity Sep 2018 to Jul 2024
    Pentester recruiting, ESXi research, EDR evasion questions and Qilin ads in the same groups
    Every identity step relies on leaked data that has not been independently verified.

    How Much Weight It Carries

    Our view is that the chain is plausible and worth tracking, but it is not proof. Several links depend on leaked Russian government, telecom and forum data whose integrity cannot be checked from the publication. The first and most important step, from Haise to the Exploit and XSS profiles, appears to rest on the order of names in approval records rather than a shared unique identifier.

    The Telegram messages fit a ransomware operator, yet interest in ESXi and EDR evasion describes a lot of people in those groups, not only an administrator. There is also a small internal inconsistency: the narrative places a Qilin ad “one week” after the July 29, 2024, post, while its own ad dates start on July 22.

    GangExposed’s track record cuts both ways. When the channel named Conti and Trickbot leader “Stern” in 2025, German police later confirmed the identity. The same reporting notes that some researchers see the leaker as a former insider settling scores or recycling older material. Either way, the channel has shown sustained access to Russian-speaking ransomware circles, and the Qilin post widens its campaign beyond the Conti ecosystem.

    Where Qilin Stands Now

    Qilin started as Agenda in July 2022, rebranded that September, and builds payloads for Windows and VMware ESXi in Rust and Go. It became one of the defining RaaS brands of 2025, adding more than 185 victims to its leak site in October 2025 alone, including claimed attacks on Asahi, US local governments and Texas power companies, per reporting at the time. Its tally now stands at more than 2,350 known organizations across 62 countries, with more than 450 victims listed since June 2026.

    The pressure is building, though. Threat group tracking shows Qilin lost the leak site volume lead to The Gentlemen in June 2026 and saw volume halve after its Q3 2026 peak, while continuing to operate after the October arrest. A public identification of its administrator adds another layer of risk for the people behind the brand, and for affiliates weighing whether to stay.

    Bottom Line

    GangExposed has put a name to one of the most visible personas in the RaaS economy, and the case it lays out is detailed enough to take seriously. It is also built almost entirely on leaked data and one investigator’s linkage, with no confirmation from the authorities who already have a Qilin suspect in custody. We treat the identification as a credible lead that still needs confirmation.

    For Qilin itself, the exposure lands at a bad moment: declining leak site volume, a suspected core member under arrest and now its public face allegedly unmasked. The more useful signal over the coming weeks will be whether its affiliates stay or move to other programs.

    FAQ

    Who Is Haise in the Qilin Ransomware Group?

    Haise is the RAMP forum persona widely described as Qilin’s administrator. The profile joined RAMP in May 2022, advertised Qilin’s affiliate program in February 2023 and later announced new panel features such as a “call lawyer” negotiation option. Researchers also used the persona to learn Qilin’s affiliate payout structure in 2023.

    Has Haise Been Officially Identified?

    No. GangExposed published an investigation naming a Russian national as Haise, but no law enforcement agency has confirmed the identification. The claim relies on leaked databases and forum records that have not been independently verified.

    Is the Qilin Member Arrested in Germany the Same Person as Haise?

    There is no public evidence of that. Germany arrested a Russian national suspected of being a leading Qilin member in October 2026 after extradition from Japan, but available reporting does not give the suspect’s alias or link them to Haise.

    Who Is GangExposed?

    GangExposed is an anonymous persona that began publishing internal Conti and Trickbot files on Telegram in May 2025 and named several alleged senior members. One of its identifications, of Conti and Trickbot leader “Stern”, was later confirmed by German police.

    Is Qilin Ransomware Still Active?

    Yes. Qilin has continued operating after the October 2026 arrest and has listed more than 450 victims on its leak site since June 2026, although its volume has fallen from its 2026 peak.

    Does Exposing a Ransomware Administrator Stop Attacks?

    Usually not immediately. Affiliates retain their access and tools, and RaaS programs can continue or rebrand. Organizations should keep securing remote access, hardening hypervisors and monitoring for credential leaks regardless of attribution news.

    What Should Defenders Do About Qilin Right Now?

    The identification does not remove affiliates’ access or tooling, so the practical picture has not changed. In many cases, Qilin affiliates have used stolen administrative credentials found on the Dark Web to get into VPNs (T1078, T1133), and the group’s ESXi payloads make hypervisors a priority target. Enforcing MFA on remote access, rotating leaked credentials and hardening ESXi management remain the most effective steps.

    blog image 4

    Source: SOCRadar · 2026-11-06

  • Dark Web Profile: Nova Ransomware

    Dark Web Profile: Nova Ransomware

    Nova is a ransomware group that operates a Ransomware-as-a-Service (RaaS) model to extort organizations through file encryption and data theft. Nova equips affiliates with a Rust-based encryptor and threatens to publish stolen information on a dedicated Dark Web leak site. With victim claims spanning multiple industries and countries, the group threatens organizations well beyond a single sector or region. The group first surfaced as RALord in March 2025 and rebranded as Nova the following month.

    Who Is Nova Ransomware (Formerly RALord)?

    Nova is a financially motivated ransomware group that recruits affiliates, supplies encryption tools, and uses stolen data to extort organizations. After the rebrand, Nova carried existing RALord victim listings onto the renamed leak site and added affiliate registration and communication channels.

    Nova RaaS threat actor card
    Nova RaaS threat actor card

    Nova claims that its members had taken part in other, unnamed ransomware groups in 2018 and 2020. Although unverified, this account suggests that the Nova brand may be newer than the individuals behind it. The group names financial gain as its primary motive and says it selects victims based on exploitable weaknesses, not industry.

    Nova advertises several ways for other threat actors to participate:

    • RaaS affiliate program: Early RALord advertisements offered affiliates 85% of ransom proceeds, and Nova advertised a 90% share in June 2025. Both figures reflect advertised terms, not verified payments.
    • Encryption-tool access: Nova offers access to its ransomware tools separately from the affiliate program.
    • Stolen-data sales: The group advertises stolen information for sale, adding another way to profit from compromised organizations.

    From RALord to Nova Ransomware

    RALord is the name under which the group first appeared in March 2025, running its own leak site and affiliate advertisements. RA World is a separate and older ransomware operation, active since 2023 after first emerging as RA Group. The shared “RA” prefix makes the two easy to confuse, which is why Nova’s own account of their relationship matters.

    The group adopted the Nova name in April 2025. In a June 2025 interview, the threat actors claimed that they had changed the name after rejecting an approach from RA World, which they said had sought to join them. This explanation remains unverified, and it does not support treating Nova as an RA World successor. Linking the two without further evidence risks conflating separate threat actors and misreading the group’s history.

    Nova uses a Rust-based encryptor, but claims linking the ransomware to leaked Babuk source code remain unconfirmed. Neither the RALord name nor Nova’s claimed contact with RA World establishes shared membership, infrastructure, or code.

    What Are the Targets of Nova Ransomware?

    Nova describes any company it considers vulnerable as a potential victim, and its leak site reflects that broad approach. SOCRadar Threat Intelligence tracked 177 organizations named on the site between April 2025 and July 2026, spanning 45 countries and 10 industry sectors. The latest posting appeared in July 2026. This distribution suggests that opportunity, not a fixed geographic or industry preference, drives Nova’s target selection.

    Top countries targeted by Nova RaaS
    Top countries targeted by Nova RaaS

    The 10 most frequently listed countries account for 55.4% of Nova’s claimed victims. The remaining 44.6% span 35 countries, a geographic reach well beyond a small cluster of national markets.

    No Commonwealth of Independent States (CIS) country appears among the top 10. However, this absence does not show that Nova avoids attacking organizations in the region.

    Top industries targeted by Nova RaaS
    Top industries targeted by Nova RaaS

    Technology and manufacturing together account for approximately 45% of Nova’s claimed victims. Government and defense organizations represent another 5.6%, showing that Nova also targets the public sector.

    Nova’s leak-site activity increased sharply in mid-2026:

    • 74 claimed victims appeared between May and July 2026.
    • Those three months account for 41.8% of all claims tracked during the collection period.

    The increase reflects a surge in public victim disclosures. Posting dates do not necessarily match when the underlying attacks occurred.

    What Are the Techniques of Nova Ransomware?

    Nova combines ransomware deployment, data theft, and public disclosure threats to extort organizations. The group’s advertisements describe how affiliates supply network access, while malware analysis documents techniques for weakening security controls, obstructing recovery, and encrypting files. Early RALord samples and later Nova builds differ, so findings from one sample do not automatically apply to every attack.

    Initial Access and Affiliate Responsibilities

    Nova asks affiliates to provide validated access to compromised networks and to extract sensitive information before ransomware deployment. The group advertises additional charges for data extraction and hosting when affiliates have not completed those steps.

    These instructions describe a division of responsibilities, not how affiliates obtain access. Available evidence does not identify a consistent Nova-specific entry vector.

    Execution and Anti-Analysis

    Nova’s analyzed Windows encryptor contains Rust standard-library paths and Cargo dependency references, which identify it as a Rust-based executable.

    Rust library paths and Cargo dependency strings embedded in a Nova ransomware sample. (SonicWall)
    Rust library paths and Cargo dependency strings embedded in a Nova ransomware sample. (SonicWall)

    [Screenshot placeholder: replace the SonicWall image with the Nova sample’s ANY.RUN or MalwareBazaar page]

    An early RALord sample examined in SonicWall’s research opened a console window and scanned files for encryption. The sample also attempted to encrypt its own executable, revealing a flaw in that build.

    A later Windows build checks for a debugger, enumerates running processes, and terminates itself when it detects specified reverse-engineering, debugging, or monitoring tools. These behaviors align with Debugger Evasion (T1622) and Process Discovery (T1057).

    Defense Evasion

    The later analyzed build uses PowerShell, registry changes, and service manipulation to interfere with endpoint protection. Nova attempts to:

    • Disable Microsoft Defender protections and add drive-wide scan exclusions.
    • Stop security services and terminate processes associated with antivirus (AV) and endpoint detection and response (EDR) products.
    • Clear Windows Event Logs across multiple channels to reduce the evidence available to investigators.

    Whether these actions succeed depends on permissions, host configuration, and protection settings.

    Nova Ransomware Development

    In an April 4, 2025 announcement for “Nova 1.0,” the threat actors claimed to have corrected weaknesses identified in published research and to have added anti-analysis and detection-evasion features. The post appeared under RALord branding, showing that the Nova name already featured in ransomware updates during the transition.

    Nova also claims to maintain five locker types, including versions for different privilege levels, a Linux variant, and a boot locker. These are advertised capabilities, not features verified across all variants.

    Data Theft and Extortion

    Nova advertises both encryption-only attacks and double extortion. Double extortion combines file encryption with data theft and hosting, letting the threat actors pressure victims through both disruption and threatened disclosure.

    The group’s instructions separate data extraction from ransomware deployment but do not name the tools, transfer protocols, or storage providers used for exfiltration. Nova’s double-extortion model therefore does not point to a specific exfiltration technique.

    Recovery Inhibition

    The analyzed Windows build enumerates logical drives through Windows Management Instrumentation (WMI) and deletes Volume Shadow Copies using vssadmin and wmic. This behavior aligns with Inhibit System Recovery (T1490) and blocks recovery through Windows shadow copies.

    Encryption and Ransom Notes

    The later analyzed Rust build uses XChaCha20-Poly1305 for file encryption and RSA-2048 to protect encryption keys. Encrypting victim files corresponds to Data Encrypted for Impact (T1486).

    File extensions and ransom-note names differ between samples:

    • Early RALord sample: Appends .ralord and creates a README text file with a randomized filename suffix. The note directs victims to negotiate through qTox without specifying a fixed ransom amount.
    • Later Nova sample: Appends .xgWLckNV and drops a ransom note named README_NOVA.me.
    Nova ransomware ransom note.
    Nova ransomware ransom note.

    The later sample also deletes its own executable after encryption. These differences make extensions and note names useful sample artifacts, but not reliable identifiers for every Nova build.

    How Can Organizations Defend Against Nova Ransomware?

    Nova’s tradecraft is opportunistic rather than novel, so the controls that blunt it are the same ones that blunt most RaaS operations. The measures below follow the tactics, techniques, and procedures (TTPs) described above, ordered by where they interrupt the intrusion chain.

    Restrict External Access Points

    Nova depends on affiliates to deliver validated network access, even though no consistent entry vector has been identified. Enforce multi-factor authentication (MFA) on all virtual private network (VPN) and Remote Desktop Protocol (RDP) access, and disable internet-facing RDP where it is not strictly required. Rotate credentials that may have appeared in credential-stuffing campaigns or initial access broker (IAB) listings.

    Protect Backup and Recovery Capability

    Nova’s analyzed builds delete Volume Shadow Copies, so shadow copies alone cannot serve as a recovery path. Maintain offline or immutable backups that the production network cannot reach, and test recovery procedures regularly. Backups do not neutralize the threat of data disclosure, but they remove ransom payment as the only path to operational recovery.

    Extend Coverage Beyond Windows Endpoints

    Nova advertises a Linux variant and a boot locker, and third-party reporting describes VMware ESXi payloads. Extend EDR coverage to Linux hosts and hypervisor management interfaces, not just the Windows workstations and servers that legacy EDR deployments often prioritize.

    Monitor for Defense Evasion Indicators

    Alert on unauthorized changes to Microsoft Defender policy, real-time protection settings, or scan exclusions, whether made through PowerShell or written directly to the registry. Alert on bulk Volume Shadow Copy deletion and mass clearing of Windows Event Logs, both of which precede encryption in the analyzed execution flow. Baseline anomalous PowerShell and command-line activity outside expected administrative windows.

    Third-party reporting also describes bcdedit commands that force Safe Mode with Networking before encryption in some Nova intrusions. Alert on unauthorized boot configuration changes made through bcdedit.

    Harden EDR Against Tampering

    Enable tamper protection on endpoint security products and alert on service-level attempts to stop or disable them. Nova’s analyzed builds terminate protection components from a range of major AV and EDR vendors rather than target a single product.

    Control Remote Access Tooling

    Maintain an allowlist of approved remote access software and alert on unauthorized installation or execution of tools such as PuTTY, AnyDesk, or LogMeIn. Third-party reporting indicates that Nova affiliates have used these tools for lateral movement and persistence.

    Integrate Dark Web and Threat Intelligence Monitoring

    Because Nova affiliates must arrive with validated access, exposed credentials and access listings are an early warning sign. Continuous monitoring of Dark Web marketplaces and leak sites for exposed corporate credentials, VPN access listings, and data leak site (DLS) postings naming your organization can shorten the time between compromise and detection. SOCRadar Threat Intelligence’s Advanced Dark Web Monitoring (ADWM) module provides this coverage. SOCRadar’s Threat Actor Intelligence module complements this by tracking Nova’s leak-site postings, claimed victims, and activity trends in a single view.

    [ Nova Ransomware profile in SOCRadar Threat Actor Intelligence
    [ Nova Ransomware profile in SOCRadar Threat Actor Intelligence

    What Are the MITRE ATT&CK TTPs of Nova Ransomware?

    The table below consolidates Nova’s techniques with their current MITRE ATT&CK identifiers. The Basis column shows whether each mapping comes from analyzed samples, from Nova’s own advertisements, or from third-party reporting.

    Tactic Technique ID
    Initial Access Valid Accounts T1078
    Initial Access Exploit Public-Facing Application T1190
    Initial Access Phishing: Spearphishing Attachment / Link T1566.001, T1566.002
    Execution Command and Scripting Interpreter: PowerShell T1059.001
    Execution Command and Scripting Interpreter: Windows Command Shell T1059.003
    Persistence Boot or Logon Autostart Execution: Registry Run Keys T1547.001
    Persistence Scheduled Task/Job: Scheduled Task T1053.005
    Defense Evasion System Binary Proxy Execution T1218
    Defense Evasion Safe Mode Boot T1562.009
    Defense Evasion Impair Defenses: Disable or Modify Tools T1562.001
    Defense Evasion Indicator Removal: Clear Windows Event Logs T1070.001
    Defense Evasion Debugger Evasion T1622
    Discovery Security Software Discovery T1518.001
    Command and Control Remote Access Tools: Remote Desktop Software T1219.002
    Lateral Movement Remote Services: SSH T1021.004
    Exfiltration Exfiltration Over Web Service T1567
    Impact Data Encrypted for Impact T1486
    Impact Inhibit System Recovery T1490

    What Are the IOCs of Nova Ransomware?

    For indicators of compromise (IOCs) associated with Nova Ransomware, visit the threat actor profile on the SOCRadar platform.

    Advanced Dark Web Monitoring

    Source: SOCRadar · 2026-11-05

  • n0n Ransomware: Inside an AI-Assisted Extortion Operation

    n0n Ransomware: Inside an AI-Assisted Extortion Operation

    In September 2026, the SOCRadar Threat Research Unit (STRU) discovered an exposed staging server holding the live workspace of an operator tied to the n0n ransomware group. The open directory exposed the operation’s full September workflow: breach and exfiltration tooling, original locker source code, extortion infrastructure, and an AI coding agent environment backed by more than $7,000 in paid LLM API usage.

    The findings implicate at least 23 organizations across 14 countries, including a US sports retailer, a cloud provider, a teachers’ union, a Venezuelan ISP, an Argentine ministry and a pharmaceutical subsidiary. No vulnerability exploitation was identified: every documented compromise relied on weak or stolen credentials. Where n0n does not encrypt, it destroys data and backups after exfiltration.

    This post summarizes the research. The full technical analysis, including locker internals, per-victim case studies and the complete operator toolkit, is available in the whitepaper, Automating Extortion: Inside an AI-Assisted n0n Ransomware Operation.

    Key Takeaways

    • Exposed workspace: An open directory at 185.157.161[.]148 exposed 1,719 files from a n0n operator’s workspace across two September 2026 snapshots, including 829 scripts and 5 C++ encryptor sources
    • Credential-only access: Entry points included a credential stuffing campaign against 305 FortiGate SSL-VPN devices, exposed AWS keys, weak RDP/WinRM credentials and compromised hosting management portals
    • Security tools turned against victims: The operator hijacked Cisco Firepower Management Center instances to switch all access policies to BLOCK, which the actor claimed caused widespread service disruption
    • AI-built tooling: An AI coding agent configured with about 150 models produced hundreds of attack scripts, with more than $7,216 in usage costs
    • Scale: As of October 19, n0n listed 93 victims with total ransom demands of $283.7M

    Who Is n0n?

    n0n is a data-exfiltration and destructive-extortion actor tracked by SOCRadar since September 2026. The group insists on the exact spelling of its name, “two lowercase n letters with a zero between them,” uses the slogan “n0n | no js | no mercy” across its Tor and clearnet portals, and appends the .n0n extension to encrypted files.

    Public leak site of the n0n ransomware group
    Public leak site of the n0n ransomware group

    The operation has scaled quickly. On October 12, 2026, the group added 38 maritime, port and logistics organizations across 18 countries to its Dark Web leak site in a single batch. Maritime security research linked at least 21 of them to customers of South Korean maritime software provider CyberLogitec, though CyberLogitec has not confirmed an incident and the supply-chain connection remains unverified. None of the 38 listings was marked as settled after the October 19 deadline, and all were published.

    Discovery and Infrastructure

    STRU found a Python HTTP file server on 185.157.161[.]148:8877 (AS42675, Sweden) exposing the operator’s working directory in two snapshots on September 25 and 26, 2026. The files map a segmented setup: an agent box hosting the AI coding agent and C2 panel, a “raas” node running the leak site backend and per-victim Bitcoin wallet generation, a “monster” server where all stolen data lands, and a VPN-bound seedbox that distributes leak torrents.

    n0n operator's infrastructure
    n0n operator’s infrastructure

    Campaign Timeline

    n0n’s tracked operations ran from September 17 to October 22, 2026. Across three publication rounds and the ChiBiTek listing on October 11, the group exposed 15 of the 23 organizations identified in the workspace, typically leaking data four days after public disclosure.

    n0n campaign timeline
    n0n campaign timeline

    Modus Operandi

    Every case follows the same lifecycle, with the main impact shifting between exfiltration, encryption, destruction and network disruption depending on the target.

    • Initial access: Credential stuffing against 305 FortiGate SSL-VPN appliances through residential proxies, plus weak RDP/WinRM credentials, exposed AWS keys and reused hosting portal sessions
    • Persistence and collection: Hidden admin accounts disguised as services, covert proxies, and theft of browser credentials, session tokens, mailboxes and emailed MFA codes
    • Exfiltration: Per-target pipelines with data budgets, sending everything to the operator’s central storage
    • Encryption and destruction: Wrapper scripts stop services, delete shadow copies and wipe free space before the locker runs; in some cases, the operator deleted data and backups instead of encrypting
    • Extortion and publication: Ransom notes, countdowns, negotiation rooms and direct extortion emails, followed by scrubbed ZIP and torrent leaks
    FortiGate gateway credential checker
    FortiGate gateway credential checker

    The n0n Encryptor

    The locker is a lightweight AES-256-CBC encryptor with Windows, Linux and PowerShell variants, plus an OpenSSL stage for firewall appliances. Most destructive steps live in external wrapper scripts. Each build uses a single hardcoded key, and one key is likely shared across two victims, so a decryptor for one could unlock the other.

    n0n execution chain
    n0n execution chain

    Extortion Playbook

    Demands range from 0.5 BTC to 33 BTC per victim, and up to $28M per organization in the October port wave. Each victim gets a unique Bitcoin address, and opening a negotiation room requires a 10% upfront payment. A check of all 55 collected addresses showed zero transactions as of October 16, 2026. To raise pressure, n0n contacts victims’ clients, MSPs, unions and the media, and in one case emailed PayPal’s security team over a breach at its contact-center provider.

    Negotiation room for FinSoft
    Negotiation room for FinSoft
    Extortion email to PayPal for third-party leverage
    Extortion email to PayPal for third-party leverage

    Case Highlights

    • Dediserve (cloud hosting): Taken over through stolen admin browser credentials, with tenant backups extracted and the only Linux locker deployment
    • Inter (Venezuelan ISP): The operator flipped every firewall rule on the ISP’s Cisco FMC to BLOCK, claimed to have severed national network traffic, and demanded 15 BTC
    • Fanatics (sports e-commerce): An exposed AWS key led to exfiltration from 40 S3 buckets and deletion of a bank-transaction archive and twelve buckets, with a 30 BTC demand
    • ChiBiTek (managed IT services): Breached end to end through the FortiGate credential stuffing pipeline

    AI-Assisted Operations

    The operator uses opencode, an open-source AI coding agent, configured with 154 model entries across seven providers, including “uncensored” and “abliterated” models with safety training removed and a self-hosted model. The agent wrote and debugged hundreds of attack scripts, and gateway spending passed $7,216 by October 22, 2026. The evidence points to a human-directed operation supported by AI; no fully autonomous attack system was observed.

    LLM key status
    LLM key status

    Victimology and Attribution

    Geographic distribution of victims
    Geographic distribution of victims

    Of the 23 organizations identified in the workspace, 15 were published on the leak site, 4 were compromised but not published, and 4 saw access attempts with no outcome shown. Victims span 15 sectors, and all demands are financial.

    STRU assesses with medium confidence that the server is a private workspace for a single operator or small team, while the group’s broader framework, tooling and master wallet control indicate a multi-operator operation. The corpus provides no definitive geographic or national attribution.

    Recommendations

    • Harden remote access: Enforce phishing-resistant MFA on RDP, SSL-VPNs and admin portals, monitor /remote/logincheck for credential stuffing, and invalidate credentials found in infostealer logs
    • Lock down firewall management: Restrict FortiManager and FMC interfaces with IP allow-lists and MFA, require approvals for policy changes, and flag REST token creation from unknown IPs
    • Reduce cloud key risk: Replace long-lived keys with short-lived roles, separate S3 list, read and delete permissions, enable Object Lock and versioning, and alert on high-volume DeleteObjects calls
    • Keep offline backups: Use air-gapped or immutable backups, since n0n purges shadow copies, wipes free space and deletes provider-hosted backups
    • Hunt now: Search EDR, CloudTrail, WinRM/RDP and firewall logs for the indicators and behaviors below

    Behavioral Detections

    • Endpoint: Creation of C:\Users\Public\n0n\ or *.n0n, *.n0n.tmp and n0n_READ_ME* files; rapid renames to .n0n after MSSQL/W3SVC/MySQL termination and vssadmin shadow deletion
    • PowerShell: Net.WebClient.DownloadFile requests to HTTPS hex URLs with ServerCertificateValidationCallback={$true}, followed by Get-FileHash SHA1 checks
    • Persistence: SYSTEM scheduled tasks referencing C:\Users\Public\n0n\ (such as [NAME]Sync*, [NAME]Audit, [NAME]Exfil, NetAssist, ChromeDbg*), Registry Run key n0nNote, and hidden local admins with non-expiring passwords and UserList=0
    • Browser: Chrome launched with –remote-debugging-port on cloned profiles, or portproxy forwarding port 9223 to 127.0.0.1:9222
    • Firewall appliances: FMC REST tokens from non-management IPs, default policy flips to BLOCK, Snort rule “n0n-lock” (sid 2100001), OpenSSL-encrypted *.n0n files under /Volume/lib, and authorized_keys entries ending in root@3090-AI
    • Cloud: CloudTrail sequences of ListBuckets, sustained GetObject and DeleteObjects in 1,000-item batches, PutObject writing n0n_READ_ME.txt, and ssm:SendCommand from new or inactive keys
    • Hosting platforms: New OnApp recipes or Zabbix scripts with “maintenance” names, backup-to-template conversions and VMs tagged mnt-*
    • Network: JSON POSTs to /x/i containing “ippay-web” with card data; FortiGate /remote/logincheck requests carrying Cookie: APBSID=n0n

    MITRE ATT&CK TTPs

    Tactic Technique Procedure
    Reconnaissance T1595.001 – Active Scanning: Scanning IP Blocks RDP batch scans and a ranked list of 305 FortiGate SSL-VPN appliances
    Initial Access T1078 – Valid Accounts Weak or default RDP/WinRM administrator credentials; firewall-manager admin accounts
    Initial Access T1078.004 – Valid Accounts: Cloud Accounts Exposed long-lived AWS IAM access keys
    Initial Access T1133 – External Remote Services Internet-facing FortiGate SSL-VPN, RDP and WinRM
    Execution T1059.001 – Command and Scripting Interpreter: PowerShell Strike, exfiltration and implant scripts pasted into RDP sessions
    Execution T1053.005 – Scheduled Task/Job: Scheduled Task Locker and audit runs as SYSTEM tasks every 5-10 minutes
    Execution T1651 – Cloud Administration Command AWS SSM commands, hosting-panel recipes and Zabbix script execution
    Persistence T1053.005 – Scheduled Task/Job: Scheduled Task AfricaSync, FinSoftSync*, SvcSync, NetAssist
    Persistence T1547.001 – Boot or Logon Autostart: Registry Run Keys n0nNote opens the ransom note at every logon
    Persistence T1136.001 – Create Account: Local Account Hidden administrators svc_upd, svc_reg, svc<slug>
    Persistence T1098.004 – Account Manipulation: SSH Authorized Keys Operator keys installed via VNC console on cloud clones; root@3090-AI on FMC appliances
    Defense Evasion T1562.001 – Impair Defenses: Disable or Modify Tools Symantec Endpoint Protection stopped and disabled
    Defense Evasion T1564.002 – Hide Artifacts: Hidden Users SpecialAccounts\UserList value 0 for the cover accounts
    Defense Evasion T1070.004 – Indicator Removal: File Deletion Toolkit overwritten three times and deleted; launchers self-delete
    Credential Access T1110.004 – Brute Force: Credential Stuffing FortiGate /remote/logincheck credential stuffing via residential proxies (APBSID=n0n)
    Credential Access T1110.002 – Brute Force: Password Cracking Offline cracking of stolen SaaS user hashes and encrypted 7z archives
    Credential Access T1555.003 – Credentials from Password Stores: Web Browsers Chrome/Edge stores decrypted, including Chrome’s app-bound key
    Credential Access T1539 – Steal Web Session Cookie Live administrator Chrome profiles copied and driven over DevTools
    Credential Access T1111 – Multi-Factor Authentication Interception Billing-panel MFA codes read from the staff mailbox
    Credential Access T1003.003 – OS Credential Dumping: NTDS NTDS.dit copy from a shadow copy (scripted)
    Discovery T1046 – Network Service Discovery Internal sweeps of 445/3389/5985/1433
    Discovery T1580 – Cloud Infrastructure Discovery EC2/ECS/SSM mapping and hosting-panel VM inventories
    Discovery T1619 – Cloud Storage Object Discovery S3 bucket listing and prefix census
    Lateral Movement T1021.001 – Remote Services: Remote Desktop Protocol Headless xfreerdp sessions driven by key injection and OCR
    Lateral Movement T1021.006 – Remote Services: Windows Remote Management pywinrm strikes, triage and browser theft
    Collection T1530 – Data from Cloud Storage S3 bulk pulls with per-bucket byte caps
    Collection T1114.002 – Email Collection: Remote Email Collection Full IMAP dump of a provider staff mailbox
    Collection T1115 – Clipboard Data RDP results returned through the clipboard
    Collection T1113 – Screen Capture Remote screen captured and read with OCR to drive logons
    Command and Control T1071.001 – Application Layer Protocol: Web Protocols HTTPS implant C2 (c2srv.py) and the kserver panel
    Command and Control T1001 – Data Obfuscation Status beacon disguised as card-payment JSON to /x/i
    Command and Control T1572 – Protocol Tunneling SSH -D SOCKS tunnels over Tor
    Command and Control T1090.003 – Proxy: Multi-hop Proxy Tor, SSH pivots, gost exits on compromised hosts, residential proxies
    Exfiltration T1041 – Exfiltration Over C2 Channel Zip-base64 POSTs to /c2/…/out (8 GiB budget)
    Exfiltration T1567.002 – Exfiltration Over Web Service: to Cloud Storage Uploads to a public file-hosting service
    Impact T1486 – Data Encrypted for Impact n0n locker (Windows/Linux) and the OpenSSL stage on FMC appliances
    Impact T1485 – Data Destruction S3 mass deletion; a provider’s servers and backups removed
    Impact T1490 – Inhibit System Recovery Shadow copies deleted; free space wiped
    Impact T1489 – Service Stop MSSQL, MySQL and IIS stopped before encryption
    Impact T1565.001 – Data Manipulation: Stored Data Manipulation Firewall rules flipped to BLOCK in the FMC configuration database

    Whitepaper

    Automating Extortion: Inside an AI-Assisted n0n Ransomware Operation

    Get the full technical analysis, including locker internals, per-victim case studies and the complete operator toolkit.


    Download the Whitepaper

    Indicators of Compromise (IOCs)

    Type Indicator Context
    IPv4 185.157.161[.]148 Agent box – kserver C2 – /x/i beacon
    IPv4 185.157.161[.]93 ntpsync[.]io origin – PHP payloads
    IPv4 81.161.239[.]10 “monster” – loot vault – implant C2 – payloads (:8455)
    IPv4 188.214.135[.]130 “raas” – leak backend – payloads (:8898, :8899)
    IPv4 198.13.158[.]143 SSH pivot over Tor – SOCKS hop to victim SaaS tenants
    IPv4 45.132.114[.]241 Operator scrub-list IP
    IPv4:port 217.18.55[.]185:11434 Self-hosted LLM
    Domain ntpsync[.]io Beacon cover domain
    Domain n0n-cases[.]net Extortion mail sender
    Email notice@n0n-cases[.]net Extortion mail sender address
    Domain n0n[.]to Leak site (clearnet, inactive)
    Domain uberip[.]com Actor mail domain
    Domain nongzecboljwv3yfndkggsybsglfrkffw7bvk2zemuteoxe6etpusnad[.]onion Onion leak site (Tor)
    File / Path *.n0n, *.n0n.tmp, n0n_READ_ME.txt, C:\Users\Public\n0n\ Encrypted files – note – staging dir
    Registry HKLM\Software\Microsoft\Windows\CurrentVersion\Run : n0nNote Ransom note at logon
    Account svc_upd (“Software Update”) – svc_reg (“Registry Service”) – svc<slug> Hidden admins
    SHA-1 dc82e9c22c4a06c6bc03701fce2b55f4f490c651 nenc.exe
    SHA-256 d1fd3a8ac4ada74b8ccbcdf5134797acff401b7fb50e16549b337781dfe1a486 nenc.cpp – Windows locker v1
    SHA-256 340172627cdf40927784956be784bd2b3f5d1584cb79cf31064f387bbe9acb28 nenc2.cpp – Windows locker v2
    SHA-256 886b3608d05069d314fa6300404de6627f96b13a95c5633f6c91dfc2a545ac87 nenc2l.cpp – Linux locker

    Bitcoin addresses (ransom wallets, September, including maritime wave wallets)

    • bc1qcf8uuvqfnucqt8crr8cgkfjevhvytjahppesem
    • bc1qxr2ruep9ast6sz32x7eza7csf49sguqwsnhhaa
    • bc1qyv92sc45cgah5ggwmsppdm323d4jya4rprrqtu
    • bc1q78uygtqyjumvh3s7dz75s0ttgsrz4jx5t3wu9n
    • bc1q98r9gusxzn4st2sr9mxvq7r66ny8dfgtdlgk9h
    • bc1q7zlg34wgtdw0dl0tn4gh52vj9nglt77p9necy8
    • bc1q97y47gkmd4cve9eadws59xz434ag2mntwy7lnv
    • bc1qezgj37g0hr8d29txsnpdjs02lr9kjvhyxzwtkg
    • bc1qp2hk3dla0tqus76emls2sppekvurrcd3ghj827
    • bc1qncxk5npnqzp39ns8jv5mj9pr6vx0ekpln2605s
    • bc1qv6mhllrmzdgl2rk09j7ynuqlgzhfguw0sj77vc
    • bc1q47wp4v8u5myse07mglenk4skssvdj6zfvdrj0l
    • bc1qh8082ye3dsyhcg8q9mglvr0w80v0gpumd5ndv8
    • bc1qpp89qpfmvf9xm6ued5v49e92s47m3mn4vngusw
    • bc1q030z6k205jz46cyn60jqts9v7xlq895ttd6r9p
    • bc1qtkuqkl5h4qekgpvjpkfgapl5fqgzpxdntlwk9c
    • bc1qck73xcn7p83emjkgrcpkvujqngnraunu09sc9s
    • bc1qthtsf88sx3czclzkx73cjxk0u5ep5kqnvhtpt8
    • bc1qp6wyngevq4aycqll3wpc03pf6qwgwcvr3md6ga
    • bc1qldy6h5zlv0aqhyyxwfsra3gqjw694yfj8rtl83
    • bc1qmfhq8plvqe4cdtypgukq7276jmerdt3gqh33xy
    • bc1qxed2puty9lmcga2mk6780khqtnlrl5ptsm6t49
    • bc1qfed3fh7f6g8dkmvq7pvrln3f0khf8aut9r2hue
    • bc1qcxqsrfrefgyv6tevuytquq28c4zrftvcflwm96
    • bc1qt2f08qa3e4wtn0rkscg4flypp3dxy7dhzefa4e
    • bc1q2ua7h64tmywa7xfn4l7f0v868tfte80lgfcmh3
    • bc1qg7nkelnag2sefg45dgu3ay7xyv3ak3v48fh45h
    • bc1qswvm6d2x8n8mpuz5x2y4wtqpy34xdxe023cw4u
    • bc1qyfc9qlsmp6jrq8y4v50ps94qtffwvzejsezsqk
    • bc1qcrm6qxz633suda3eacqxkdnx05ztd2mh0y3dxa
    • bc1qxktcy6ssugttfr6v5695kzgud9eqw8h7d2r0ss
    • bc1qe9lq4m7rjgk533u3g3cgf7xs5pn2l8e3l5pfjt
    • bc1q48vuydq5j8e98lwmt69aqhr5gcu5mn2pc4eqkx
    • bc1q8etz0apvr6tclxxznsss2tk07v46ckxmrjwsde
    • bc1qpf72d8lk4caqs3ewej75jasd44umg3juw58asn
    • bc1qhavzrsz8y2lqcwp5frckl5fy8exhmacyxz3lcy
    • bc1qqrcegl2a2tkrdln4r6jn2pm9fp84mv60d5y4ug
    • bc1qsv8yzm09njxmt5khlpyrhnjfupzpd7vykc7e46
    • bc1qpek2637ed7a4f2nff60rrng52fudqsk9l905ra
    • bc1q3nswa35s5vax3q3txf3mpy260zad4vyadpfg98
    • bc1ql0pdlaga5q3xrznd80mzuzgsv4nuz4fzut7td6
    • bc1qqrvt0ank3y2axmuzxlm2z4879qqxa37nwdafvr
    • bc1qsclrrwmjnfxzwcmskqlax9htdnc9yd5n99l5ml
    • bc1qk9s5ngavpwla07w2v3wjwwkt75uj6zmspkxg6p
    • bc1qx3u3xa7eqkukxuuyr00zq5ynx5ylepw9e4d4ph
    • bc1qccgses0q46lpttjtuchnsc2rz28qn9lng67hne
    • bc1qqc2dy605g0hmcaqfm8ne85dwctfw7603f7a03m
    • bc1qmdzpn0d0flf4pswu8ze57e7yyx883lxp55f7f2
    • bc1qpjvl29njpedxhljq0nvvskmxra8p9e8stssegm
    • bc1q50x7akwznysr8r3v00p8uzxqztuzg5t2x0tavt
    • bc1qmz68h5vyqa9feduhceh34ygmaq3fr3xnau7myf
    • bc1q3jgtpksgg5ldhq34z93xqtt39y9kzpkz2674x8
    • bc1qgnmfhcppgdynjfwxzrrspxsfr9zgps6ctgkh2z
    • bc1q3aw0gz7nnqvfnv9nld686q05lgh2d4yxmzlgcg
    • bc1q9k3lqn4zfl733pxcwk7erqtdmuw68w80hvru39

    For the full IOC list, consult the SOCRadar platform.

    FAQ

    What Is n0n Ransomware?

    n0n is a data-exfiltration and destructive-extortion ransomware group active since September 2026. It encrypts files with the .n0n extension, destroys data and backups where it does not encrypt, and runs a Tor leak site with keyed negotiation rooms. As of October 19, 2026, it listed 93 victims with total demands of $283.7M.

    How Does n0n Gain Initial Access?

    Every documented n0n intrusion relied on weak or compromised credentials, and no exploit payloads or CVE references were found in the operator’s files. Entry points include credential stuffing against FortiGate SSL-VPN portals through residential proxies, weak RDP/WinRM credentials, exposed AWS IAM keys and reused hosting provider sessions.

    How Does n0n Use AI?

    The operator uses the opencode AI coding agent, configured with 154 model entries across seven providers, including uncensored, abliterated and self-hosted models. The agent produced and debugged hundreds of attack scripts, with gateway spending above $7,216. The evidence shows a human-directed operation supported by AI; no fully autonomous attack system was observed.

    Can Files Encrypted by n0n Be Recovered?

    Each n0n build uses a single hardcoded AES-256 key with no per-victim derivation, and one key is likely reused across two victims, which creates recovery opportunities for affected organizations. Because the operator also deletes shadow copies, wipes free space and removes provider-hosted backups, offline or immutable backups remain the only reliable recovery path.

    Advanced Dark Web Monitoring

    Source: SOCRadar · 2026-10-28

  • Insider Threat Report: Dark Web Recruitment & Access Trends

    Insider Threat Report: Dark Web Recruitment & Access Trends

    Blog

    Insider Threat Report: Dark Web Recruitment & Access Trends

    Flashpoint’s monthly analysis of insider threat recruitment, illicit access advertising, and threat actor activity targeting enterprise environments.

    Default Author Image
    October 15, 2026

    Unknowingly, a member of key personnel is living two separate lives. On the clock, they are a highly-trusted systems administrator, but in their personal time, they moonlight on the deep and dark web, advertising their trust and access to the highest bidder. One day, they get a simple offer: $15,000 in the crypto of their choice to approve a single push notification at 2 AM. They accept. By morning, the attacker walks away  with active domain admin credentials without the need for malware or cracking firewalls.

    This is just one example of how insider threats lead to modern enterprise breaches. As perimeter security, EDR coverage, and other security tools mature, threat actors continue to target the human element, seeking employees and contractors willing to provide access to corporate systems or facilitate fraudulent activity.

    In a threat landscape where identity is becoming the primary attack surface, monitoring illicit marketplaces and recruitment efforts is critical. This new monthly report leverages Flashpoint’s Primary Source Collection (PSC) to analyze insider threat tactics, tracking active recruitment and advertising on dark web forums and encrypted networks.

    The Insider Threat Landscape: September 2026

    In September 2026, Flashpoint analysts identified a total of 9,787 insider-related posts across 1,216 channels and 1,778 authors. While the volume of observed activity decreased, threat actors continued to actively seek and advertise insider access, particularly within industries where legitimate employee privileges can be exploited to facilitate fraud. These communications included threat actors attempting to recruit insiders at target organizations, insiders advertising their services, and general discussions involving insider-related threat activity.

    Of these communications, Flashpoint observed 857 unique posts between September 1 and September 30, 2026.

    The number of unique insider posts collected over the past 12 months. (Source: Flashpoint)

    The majority of these posts appeared in mid-tier English-language Telegram channels, where threat actors sought or advertised insider access to organizations across several industries. Much of the observed activity focused on recruiting insiders to facilitate fraud operations.

    Where Insider Threat Activity is Concentrated

    Retail, telecommunications, and financial services remained prominent targets for insider threat activity in September 2026. However, a significant portion of the observed activity could not be attributed to a specific industry.

    Flashpoint found that 45.7% of unique insider posts fell into the “Unknown” category, meaning analysts could not determine the targeted organization or industry. Among posts with an identifiable industry, retail accounted for the largest share, followed by telecommunications, financial services, and technology.

    Together, these four industries accounted for 419 unique posts, or nearly 49% of all unique insider-related activity observed during the month.

    The concentration of activity in retail, telecommunications, and financial services reflects continued threat actor interest in insiders who may be positioned to facilitate fraud, provide access to customer accounts, or exploit legitimate organizational privileges.

    Unique insider posts broken down by industry in September. (Source: Flashpoint)

    The following table shows a breakdown of unique insider posts by industry in September 2026:

    Industry Posts
    Other 392
    Retail 135
    Telecom 116
    Financial 85
    Technology 83
    Public Sector 41
    Healthcare 3
    Media 2
    Total 857

    Insider Threats: Recruiting vs. Advertising

    Active insider threats generally involve two types of activity: an insider is recruited by a malicious outside party, or an insider advertises their access and services to interested threat actors. These arrangements can enable adversaries to exfiltrate sensitive data, install malware, compromise accounts, or carry out fraudulent activities such as SIM swaps.

    In September 2026, Flashpoint analysts identified 15 notable insider recruitment and advertising posts across underground communities and chat forums. Of these, nine involved threat actors attempting to recruit insiders, while six involved actors advertising insider services.

    Telecommunications accounted for the largest number of notable posts, with five instances of recruitment or advertising. Retail and technology followed with three each, while financial services and the public sector accounted for two each.

    These examples illustrate how threat actors continue to seek access across industries, with particular interest in employees and contractors who can facilitate fraud or provide privileged access to organizational systems.

    Flashpoint also observed numerous insider solicitations repeated across multiple discussion threads, often with limited details about the access or services being offered. Analysts could not independently verify the legitimacy of the recruitment and advertising claims.

    Unique insider posts broken down by industry and whether the insider was recruited or advertised in September. (Source: Flashpoint)

    Protect Against Insider Threats Using Flashpoint

    Insider threats are inherently difficult to detect using internal security controls alone because the malicious activity relies on valid credentials and legitimate access privileges. Relying solely on internal logs means security teams often only detect an insider threat after data exfiltration or system sabotage has already occurred.

    Flashpoint protects organizations against insider threats through our Primary Source Collection (PSC) and specialized intelligence platforms:

    • External Threat Intelligence & Early Warning: Flashpoint monitors deep and dark web forums, invite-only threat communities, and encrypted chat platforms to identify employee solicitations, stolen corporate domain mentions, and active recruitment attempts before an intrusion develops.
    • Identity Protection & Infostealer Tracking: By tracking illicit marketplaces and infostealer activity, Flashpoint identifies compromised corporate credentials and active session tokens, preventing threat actors from utilizing purchased access.
    • User & Entity Behavior Context: Flashpoint’s intelligence equips SOC, Security Operations, and Risk Management teams with adversary TTPs, enabling security operations to look for anomalous data downloads, off-hours access, or unauthorized software installation.

    To learn more about how Flashpoint can help protect your enterprise from insider risk and monitor illicit underground communities, Request a Demo Today.

    Frequently Asked Questions (FAQs)

    What is the Flashpoint Insider Threat Report?

    The Flashpoint Insider Threat Report is a monthly intelligence brief that analyzes trends, volume, targeted industries, and tactics surrounding insider threat recruitment and illicit access advertising on the deep web, dark web, and encrypted chat channels.

    How does Flashpoint collect insider threat data?

    Flashpoint collects data using its Primary Source Collection (PSC) engine, which actively monitors thousands of dark web forums, illicit marketplaces, and underground chat networks where threat actors and malicious insiders communicate.

    What is the difference between insider recruitment and insider advertising?

    Insider recruitment occurs when an external cybercriminal attempts to entice a corporate employee into assisting with a cyberattack. Insider advertising occurs when an employee or contractor proactively lists their legitimate access or services for sale on illicit marketplaces.

    See Flashpoint in Action

    Source: Flashpoint · 2026-10-15