A data-theft and destructive-extortion group that surfaced in September 2026 and built its reputation on an AI-assisted operation and a leak site listing hundreds of victims — while its own exposed server showed a far smaller operation. Tracked here: history, attack path, behaviour and procedures, indicators and victimology, with every claim tied to its source and every leak-site listing labelled as a claim until a victim confirms it or data is published.
Credential-only access — no vulnerability exploitation has been documented in any intrusion. The group exfiltrates data, then encrypts files with the .n0n extension or destroys data and backups instead, and pressures victims through a Tor leak site.
SOCRadar, from the group’s own exposed workspace, October 2026.
Bitdefender: the site listed more than 100 victims, but data from the group’s own file server shows it targeted just over 20 organisations. Treat every listing as a claim, not a confirmed breach.
Bitdefender Threat Debrief, November 2026.
The profile in one table. Spelling matters: the group insists on its name as two lowercase n letters with a zero between them.
| Attribute | Detail |
|---|---|
| Name | n0n — two lowercase n letters with a zero between them; written N0n in some reporting |
| Type | Financially motivated data-theft and destructive-extortion group |
| First seen | 18 September 2026 (CyberXTron); leak site listing victims within four days |
| Slogan | “n0n | no js | no mercy” — carried across its Tor and clearnet portals |
| Leak site | Tor-hosted data leak site; clearnet portal n0n.to (inactive); RSS feed; active / leaked / honor listings; keyed negotiation rooms |
| Encryptor | Lightweight AES-256-CBC locker, Windows / Linux / PowerShell variants plus an OpenSSL stage for firewall appliances; appends .n0n; ransom note n0n_READ_ME.txt |
| Initial access | Credential-only: infostealer-harvested credentials, credential stuffing against FortiGate SSL-VPN, weak RDP/WinRM credentials, exposed AWS keys, reused hosting-portal sessions. No CVE has been tied to an intrusion. |
| Extortion model | Double extortion plus a declared threat to encrypt or destroy backups and shadow copies; demands 0.5–33 BTC per victim, up to $28M in the maritime wave; negotiation room requires a 10% upfront payment |
| Attribution | None established. Some vendors list the group as likely Russia-based without showing evidence; SOCRadar found no definitive geographic or national attribution. Evidence cited by Cydome indicates the group is not related to Iran, China or North Korea. |
| Assessment | Opportunistic and broad, with a maritime / logistics concentration in October and a US and South Korea focus. Bitdefender has high confidence n0n does not have the means to sustain a top-10 ransomware ranking. |
Newest first. Each entry is dated and sourced; leak-site claims are labelled as claims.
Ekwiik (Ideal Prepaid Botswana) de-anonymised. The leak-site listing carried anonymously as “Company #216” since 13 November is named as Ekwiik, a prepaid utilities and airtime platform in Botswana. Status [NOT PAID — EXPOSED]. No data volume, data types, ransom amount or screenshots are disclosed, and no victim statement has been reported; the claim is unverified.
Ransomware.live, discovered 20 Nov 2026 · RedPacket Security, 20 Nov 2026
Correction: Alo Yoga’s data was published on 14 November. Ransomware.live records the Alo Yoga listing as discovered on 14 November at 22:08 UTC, with the data described on the leak site: 209,051 distinct customer email addresses captured from the company’s order and returns links, with order numbers, plus shopper search sessions, per-query revenue and conversion intelligence, and site-navigation revenue graphs running to tens of millions of rows. The listing was first carried anonymously as “Company #144” on 7 November and carries [NOT PAID — EXPOSED] status. This corrects the 19 November entry below, which stated that no publication of any of the nine retail-batch datasets had been reported: Alo Yoga’s data was already published before the batch was named on 15 November. The 19 November entry is left as written. The other eight datasets in the batch remain unpublished as far as reported, and the claim itself is unverified.
Ransomware.live, discovered 14 Nov 2026 · QPulse, 15 Nov 2026
Retail-batch deadline passes with no publication reported. The countdown on the nine-organisation retail batch named on 15 November — Samsung, Starbucks, Gap, VF Corporation, Timberland, Alo Yoga, Pollstar, Steeden Financial and Airport Appliance — expired on 19 November; as of 01:00 UTC on 20 November no publication of any of the nine datasets, sample release, delisting or extension had been reported. Samsung’s data was published separately on 19 November (220MB, about 150,000 smart TV viewing events) and is covered in the next entry. The listings still carry their “NOT PAID — EXPOSED” status, which is the status they carried at discovery and does not by itself evidence a post-deadline publication. Absence of reporting is not proof nothing was published.
Visa, Mastercard and JPMorgan data published; JPMorgan confirms a third-party vendor attack. After the 48-hour deadlines expired, n0n published the three datasets on 19 November. Cybernews researchers examined the dumps: more than 6,000 premium Visa customers affected, the data likely from a marketing platform for Visa’s premium dining programmes in Asia-Pacific; roughly 16,000 JPMorgan customer email addresses from the invitation-only Reserve programme, with phone numbers in some cases, account IDs and 124 employee email addresses; and about 2,000 Mastercard entries in a 280KB file that researchers say looks like test data rather than genuine customer information, its origin and authenticity unverified. Cybernews found similar files and formatting across the three datasets, suggesting a single third-party source rather than direct breaches of the three firms. JPMorgan confirmed the same day that third parties providing services to the firm were the target of a cyberattack and that attackers accessed certain marketing data held by those vendors, which may include some customer names and email addresses; the bank said services and sensitive financial or payment data were unaffected, and that the attack is not connected to the service disruption earlier in the week. Visa’s position is unchanged: its systems were not involved and no card numbers were affected. The same day n0n also published data claimed from Samsung (220MB, including about 150,000 smart TV viewing events) and a limited Walmart sample (about 78 employee names and email addresses, with no evidence Sam’s Club member data was accessed). Starbucks had not been published as of 15:22 UTC, with its deadline still running.
Cybernews, 19 Nov 2026 (updated 12:25 UTC) · Cybernews — Mastercard, 19 Nov 2026 · Cybernews — JPMorgan, 19 Nov 2026 · Reuters, 19 Nov 2026 · Cybernews — Samsung, 19 Nov 2026 · Cybernews — Walmart, 19 Nov 2026
GRAIL, Inc. named. The US cancer-screening diagnostics company, listed anonymously as “Company #204” since 9 November, is named with claimed healthcare provider data: 56,684 provider records with NPI numbers, names, phones, addresses and Medicare payment figures, plus CRM contact records with personal and corporate email addresses and customer identification records. Status [NOT PAID — EXPOSED]. No victim statement and no samples reported; the claim is unverified, and unlike the Visa, Mastercard and JPMorgan listings it carries no Qubole linkage note.
Ransomware.live, discovered 19 Nov 2026 · Undercode News, 19 Nov 2026
The 16 November batch is most likely one supplier breach, not 59 separate ones. Ransomware.live adds the same notice to the Visa, Mastercard, JPMorgan Chase, HappyFresh, Azurity, Servier, Vir Biotechnology and Noon Academy listings: each is “most likely linked to a third-party breach of Qubole (a data platform provider), rather than a direct compromise of the victim’s own systems.” The underlying campaign was documented by Wiz: between 14 and 26 October an attacker used IAM credentials from Qubole’s AWS account to assume customer integration roles and steal secrets and S3 data. Wiz’s 11 November update relays a report that an actor calling itself n0n claimed responsibility; Wiz has not verified that claim, and Wiz says it has identified no ties between the campaign and other publicly reported activity. Expedia confirmed a supplier breach on 12 November but has not named Qubole, and Qubole and its owner Idera have published no statement. The Visa, Mastercard and JPMorgan rows below are therefore most likely supplier-held marketing data, not direct breaches — but the link is assessed, not confirmed.
Wiz, 9 Nov 2026 (updated 11 Nov) · The Hacker News, 9 Nov 2026
Three more named companies respond; Jazz Pharmaceuticals named. Eli Lilly says its internal investigation found no evidence of unauthorised access to its own systems; the data was held by a third-party vendor supplying anonymised records, and the records are not Lilly’s own patients’. Vertex Pharmaceuticals says it is investigating a potential incident involving an external data provider and that no proprietary company data or patient information was exposed. Vir Biotechnology confirms awareness, says it activated its cybersecurity procedures, and found no indication its systems were accessed without authorisation. The vendor in the Lilly case has not been publicly identified. Separately, Jazz Pharmaceuticals is named (listed anonymously as “Company #206” since 9 November) with claimed specialty-pharmacy sales records of more than 1.5 million rows carrying patient identifiers, a customer file of more than 10 million rows and a healthcare-professional directory of more than 1.2 million rows. No victim statement and no samples reported; the claim is unverified, and unlike the Visa, Mastercard and JPMorgan listings it carries no Qubole linkage note.
BeInSure, 18 Nov 2026 (reporting Barron’s) · RedPacket Security, 18 Nov 2026
Visa, Mastercard and JPMorgan deadline passes with no publication reported. The 48-hour deadline on the three listings expired on 18 November; as of 19 November no publication of their data, sample release, delisting or extension had been reported. The listings still carry their “NOT PAID — EXPOSED” status, which is the status they carried at discovery on 16 November and does not by itself evidence a post-deadline publication. Absence of reporting is not proof nothing was published.
59 organisations named in a single day, including Eli Lilly, Visa, Mastercard and JPMorgan Chase. 46 of the 59 had sat on the site since 9 November as numbered placeholders (“Company #150” and similar); n0n swapped the numbers for names and marked all 59 unpaid. Visa confirmed an incident at a third-party provider in Asia Pacific and said its own systems were not involved and no card numbers were affected; JPMorgan Chase said its 16 November outage was a software bug with no evidence of a cyberattack. The Ransomtracker holds 363 n0n entries, 168 of them numbered placeholders. All 59 are claims, not confirmed breaches.
MediaMath listed (unverified claim). Detection content published for the group’s FortiGate SSL-VPN credential-stuffing pipeline: bursts of failed logins at /remote/login, one source address trying many distinct usernames within 10 minutes, run through a 15-thread residential SOCKS5 proxy pool. No CVE is involved; access relies on weak or stolen credentials.
Bitdefender’s Threat Debrief exposes the inflation. October data-leak-site claims totalled 1,072 victims; excluding 126 n0n entries the total falls to 946. The group’s exposed file server shows it targeted just over 20 organisations, against 100+ listings. Bitdefender: high confidence n0n cannot sustain a top-10 ranking. Check Point Research, the same week: a record 1,063 ransomware attacks in October (+35% year on year); n0n third with 8% of published attacks, up from about 2% in September, with 170+ claimed victims by early November.
Bitdefender, 10 Nov 2026 · Check Point Research, 11 Nov 2026
About 189 victims claimed since mid-September. The group ranks among the top three ransomware operators alongside The Gentlemen and Qilin. Primary geographic focus is the US and South Korea; most-targeted sectors in descending order are logistics and shipping, technology, retail, healthcare, manufacturing and finance. Persistence observed: Windows scheduled tasks with a *sync naming convention, rogue administrator accounts with an svc_* naming convention, and SSH authorized keys on cloud hosts. Initial access primarily credential stuffing against Fortinet, RDP and WinRM.
The operator’s workspace is exposed. SOCRadar finds a Python HTTP file server at 185.157.161[.]148:8877 exposing 1,719 files across two September snapshots — 829 scripts, five C++ encryptor sources, extortion infrastructure and an AI coding-agent environment. The findings implicate at least 23 organisations across 14 countries. No vulnerability exploitation was identified: every documented compromise relied on weak or stolen credentials. The Register reports the exposure on 30 October.
Maritime deadline passes; none of the 38 settled and all were published. The group lists 93 victims with total ransom demands of $283.7M.
38 maritime, port and logistics organisations listed in a single batch across 18 countries, with a common publication deadline of 19 October, 13:27 UTC, and demands from $400,000 to $28 million. Maritime security research traced at least 21 of the listings to customers of South Korean maritime software provider CyberLogitec; CyberLogitec has not confirmed an incident and the supply-chain connection remains unverified. The small, application-scoped data sets point to one shared system rather than 38 separate intrusions.
21 victims listed before the maritime wave, spanning investment firms, telecommunications, retail, education and pharmaceuticals across 17 countries, including claims against Fanatics, Transcom WorldWide, AstraZeneca Türkiye, Argentina’s Ministry of Education, STOKR and Venezuelan ISP Inter.
Fortra publishes the first vendor explainer. The group is described as a data-theft extortion operation using valid credentials harvested by infostealer malware, legitimate admin tools and RDP — not an exotic vulnerability. Early headline claims relayed: Transcom (86.7 million support-agent records claimed from PayPal’s Citrix and authentication systems), Inter (15.3 million subscriber connection records claimed), Fanatics (~108 GB claimed), TapClicks (full source code claimed) and Dediserve (46 servers “paused”, 167 customer accounts claimed). All unconfirmed.
13 victims across 10 countries within four days of surfacing. United States (3), Vietnam (2), Uzbekistan (1), Mali (1), Venezuela (1), Sweden (1), Türkiye (1), Argentina (1), Luxembourg (1), Brazil (1). Financial services is the largest share at 23% (3 victims); technology, retail & e-commerce and education each 15%. 11 of the 13 listings had already passed their deadlines and had data leaked; 2 remained active.
n0n first surfaces. The group appears with a Tor leak site under the tagline “no js | no mercy” and, within four days, lists more than a dozen claimed victims. The speed suggests an already structured operation at discovery, not a group in its embryonic phase.
One lifecycle, with the impact shifting between exfiltration, encryption, destruction and network disruption depending on the target. Reconstructed from the exposed operator workspace (SOCRadar).
Credential stuffing against 305 FortiGate SSL-VPN appliances through residential proxies (requests to /remote/logincheck carrying Cookie: APBSID=n0n), weak or default RDP/WinRM administrator credentials, exposed long-lived AWS IAM access keys, and reused hosting-provider portal sessions. The earliest reporting attributes the credentials to third-party infostealer malware. No exploit payloads and no CVE references were found in the operator’s files.
Maritime-wave variant: for the 12 October batch, the evidence points the other way — small, application-scoped data sets and incomplete infostealer coverage across the victims suggest an exposed shared application or vulnerability (a terminal system, EDI integration or file-transfer product), not 38 separate credential intrusions. The shared source is unconfirmed.
Hidden administrator accounts disguised as services (svc_upd, svc_reg, svc<slug>), scheduled tasks named *Sync, NetAssist and similar, covert proxies, and SSH authorized keys on cloud hosts. The operator steals browser credentials (including Chrome’s app-bound key), live session cookies via cloned Chrome profiles driven over DevTools, mailboxes (a full IMAP dump of a provider staff mailbox), and emailed MFA codes. Discovery sweeps cover 445/3389/5985/1433 internally, plus EC2/ECS/SSM mapping and S3 bucket listing.
Per-target pipelines with data budgets, sending everything to the operator’s central storage (the “monster” server). S3 bulk pulls run with per-bucket byte caps; exfiltration uses zip-base64 POSTs to the C2 channel (8 GiB budget) and uploads to a public file-hosting service. Data typically leaks about four days after public disclosure.
Wrapper scripts stop services (MSSQL, SQL Agent, SSAS, SSIS, SQLWriter, W3SVC, MySQL), delete shadow copies and wipe free space before the locker runs. The locker is AES-256-CBC with a single hardcoded key per build. In some cases the operator deletes data and backups instead of encrypting — S3 mass deletion in 1,000-item batches, and a hosting provider’s servers and backups removed. On Cisco Firepower Management Center appliances, the operator flips firewall rules to BLOCK and injects a Snort rule (n0n-lock, sid 2100001); in the Inter case the group claimed this severed national network traffic.
Ransom notes, countdowns, keyed negotiation rooms (opening one requires a 10% upfront payment) and direct extortion emails. To raise pressure, n0n contacts victims’ clients, MSPs, unions and the media — in one case emailing PayPal’s security team over a breach at its contact-centre provider. Leaks are published as scrubbed ZIP archives and torrents distributed from a VPN-bound seedbox. Each victim gets a unique Bitcoin address; a check of all 55 collected addresses showed zero transactions as of 16 October 2026.
The group’s behaviour, mapped to MITRE ATT&CK. Select a technique to see the procedure n0n actually used, how it was observed, and what to hunt for. The full procedure table follows the diagram, so the content stands even without scripting.
Notes on the mapping. The confirmed procedures come from SOCRadar’s analysis of the operator’s exposed workspace (28 Oct 2026). CyberXTron’s first mapping (22 Sep 2026) lists techniques as “associated” without published procedures — those are marked Reported. Threadlinqs detection content adds parent techniques (T1090, T1114, T1567) and two Azazel-linked items (T1036.004, T1048) — marked Aggregator-reported and not adopted as n0n procedures; see Possible connections. One correction: CyberXTron’s early table lists “T1685 — Disable or Modify Tools”; no such technique exists in MITRE ATT&CK. The behaviour matches T1562.001, confirmed in SOCRadar’s mapping, and T1685 is not carried as a technique here.
Every technique in the diagram, with the procedure observed. This table is the reference; the diagram above is the navigation.
| Tactic | Technique | Procedure observed | Confidence |
|---|---|---|---|
| Reconnaissance | T1595.001 | RDP batch scans; a ranked list of 305 FortiGate SSL-VPN appliances. | Confirmed |
| Initial Access | T1078 | Weak or default RDP/WinRM administrator credentials; firewall-manager admin accounts. | Confirmed |
| Initial Access | T1078.004 | Exposed long-lived AWS IAM access keys. | Confirmed |
| Initial Access | T1133 | Internet-facing FortiGate SSL-VPN, RDP and WinRM. | Confirmed |
| Execution | T1059.001 | Strike, exfiltration and implant scripts pasted into RDP sessions. | Confirmed |
| Execution / Persistence | T1053.005 | Locker and audit runs as SYSTEM tasks every 5–10 minutes; persistence tasks named AfricaSync, FinSoftSync*, SvcSync, NetAssist. | Confirmed |
| Execution | T1651 | AWS SSM commands, hosting-panel recipes and Zabbix script execution. | Confirmed |
| Persistence | T1547.001 | Registry Run key n0nNote opens the ransom note at every logon. | Confirmed |
| Persistence | T1136.001 | Hidden administrators svc_upd (“Software Update”), svc_reg (“Registry Service”), svc<slug>. | Confirmed |
| Persistence | T1098.004 | Operator keys installed via VNC console on cloud clones; root@3090-AI on FMC appliances. | Confirmed |
| Defense Evasion | T1562.001 | Symantec Endpoint Protection stopped and disabled. | Confirmed |
| Defense Evasion | T1564.002 | SpecialAccounts\UserList value 0 for the cover accounts. | Confirmed |
| Defense Evasion | T1070.004 | Toolkit overwritten three times and deleted; launchers self-delete. | Confirmed |
| Defense Evasion | T1027 | Obfuscated files or information — listed as associated in the early mapping; no procedure published. | Reported |
| Defense Evasion | T1036.004 | Masquerading — reported by Threadlinqs for the Azazel-linked cluster; not in SOCRadar’s n0n mapping. | Aggregator |
| Credential Access | T1110.004 | FortiGate /remote/logincheck credential stuffing via residential proxies (Cookie: APBSID=n0n); 15-thread SOCKS5 pool per Threadlinqs. | Confirmed |
| Credential Access | T1110.002 | Offline cracking of stolen SaaS user hashes and encrypted 7z archives. | Confirmed |
| Credential Access | T1555.003 | Chrome/Edge credential stores decrypted, including Chrome’s app-bound key. | Confirmed |
| Credential Access | T1539 | Live administrator Chrome profiles copied and driven over DevTools. | Confirmed |
| Credential Access | T1111 | Billing-panel MFA codes read from the staff mailbox. | Confirmed |
| Credential Access | T1003.003 | NTDS.dit copy from a shadow copy (scripted). | Confirmed |
| Discovery | T1046 | Internal sweeps of 445/3389/5985/1433. | Confirmed |
| Discovery | T1580 | EC2/ECS/SSM mapping and hosting-panel VM inventories. | Confirmed |
| Discovery | T1619 | S3 bucket listing and prefix census. | Confirmed |
| Lateral Movement | T1021.001 | Headless xfreerdp sessions driven by key injection and OCR. | Confirmed |
| Lateral Movement | T1021.006 | pywinrm strikes, triage and browser theft. | Confirmed |
| Collection | T1530 | S3 bulk pulls with per-bucket byte caps. | Confirmed |
| Collection | T1114.002 | Full IMAP dump of a provider staff mailbox. | Confirmed |
| Collection | T1115 | RDP results returned through the clipboard. | Confirmed |
| Collection | T1113 | Remote screen captured and read with OCR to drive logons. | Confirmed |
| Collection | T1074 | Data staged — listed in the early mapping; per-target staging pipelines with data budgets in the workspace account. | Reported |
| Collection | T1114 | Email collection — parent technique reported by Threadlinqs; SOCRadar’s T1114.002 is the confirmed sub-technique. | Aggregator |
| Command & Control | T1071.001 | HTTPS implant C2 (c2srv.py) and the kserver panel. | Confirmed |
| Command & Control | T1001 | Status beacon disguised as card-payment JSON to /x/i. | Confirmed |
| Command & Control | T1572 | SSH -D SOCKS tunnels over Tor. | Confirmed |
| Command & Control | T1090.003 | Tor, SSH pivots, gost exits on compromised hosts, residential proxies. | Confirmed |
| Command & Control | T1090 | Proxy — parent technique reported by Threadlinqs; SOCRadar’s T1090.003 is the confirmed sub-technique. | Aggregator |
| Exfiltration | T1041 | Zip-base64 POSTs to /c2/…/out (8 GiB budget). | Confirmed |
| Exfiltration | T1567.002 | Uploads to a public file-hosting service. | Confirmed |
| Exfiltration | T1567 | Exfiltration over web service — parent technique reported by Threadlinqs; SOCRadar’s T1567.002 is the confirmed sub-technique. | Aggregator |
| Exfiltration | T1048 | Exfiltration over alternative protocol — reported by Threadlinqs for the Azazel-linked cluster; not in SOCRadar’s n0n mapping. | Aggregator |
| Impact | T1486 | n0n locker (Windows/Linux) and the OpenSSL stage on FMC appliances; files renamed to .n0n. | Confirmed |
| Impact | T1485 | S3 mass deletion; a provider’s servers and backups removed. | Confirmed |
| Impact | T1490 | Shadow copies deleted; free space wiped. | Confirmed |
| Impact | T1489 | MSSQL, MySQL and IIS stopped before encryption. | Confirmed |
| Impact | T1565.001 | Firewall rules flipped to BLOCK in the FMC configuration database. | Confirmed |
Technique names link to MITRE ATT&CK. “Confirmed” means the procedure was observed in the operator’s exposed workspace and published by SOCRadar. “Reported” means the technique appears in an early mapping without a published procedure. “Aggregator” means it appears in third-party detection content only.
Newest first. This is the section that moves: when a new technique is observed, a procedure changes, or a reported technique is confirmed or dropped, it goes here at the top with a date. Techniques are never silently removed — a removal gets a dated entry too.
Initial-access understanding splits for the 16 November batch. The batch most likely traces to one supplier breach, not 59 separate compromises: stolen vendor IAM credentials used for cross-account AssumeRole against Qubole customer roles (Wiz, 9 Nov; ransomware.live notices, 18 Nov). That is Valid Accounts: Cloud Accounts (T1078.004, already confirmed), not the credential-stuffing pipeline used elsewhere. No new technique is added; the n0n link to the Qubole campaign is reported, not verified, and the shared source for the batch is assessed, not confirmed.
FortiGate stuffing pipeline detailed. Detection content for the credential-stuffing campaign specifies the pattern: bursts of failed FortiGate SSL-VPN logins at /remote/login, one source address trying many distinct usernames within 10 minutes, run through a 15-thread residential SOCKS5 proxy pool. This refines T1110.004 (already confirmed) — no new technique, but the detection procedure is now specific enough to hunt on.
Azazel / MCP association reported — not adopted. Aggregator detection content links n0n to the Azazel MCP reverse-shell C2 cluster (Threadlinqs TL-2026-5634, attribution “low confidence”). The Azazel operation is documented separately by CloudSEK as a Gentlemen RaaS affiliate running its own LEAKNED site with a Penelope reverse shell exposed to an AI coding assistant through an MCP server. The two exposed workspaces are different servers with different toolsets, and SOCRadar’s n0n mapping contains no MCP technique. T1036.004 and T1048 stay marked Aggregator-reported, and the link is tracked under Possible connections, not as n0n TTPs.
Full procedure set lands at once. SOCRadar’s analysis of the exposed operator workspace replaces the early CyberXTron table as the authoritative mapping. Newly confirmed: reconnaissance (T1595.001), cloud techniques (T1078.004, T1580, T1619, T1651, T1530), MFA interception (T1111), NTDS dumping (T1003.003), browser credential and session theft (T1555.003, T1539), command-and-control (T1071.001, T1001, T1572, T1090.003), exfiltration over C2 (T1041, T1567.002), and firewall manipulation (T1565.001). The workspace also confirms the locker is real (T1486) while showing destruction (T1485) is used where encryption is not — settling the earlier disagreement over whether n0n is “really ransomware”.
Initial-access understanding splits for the maritime wave. The 38-organisation batch does not fit the credential-access pattern: SOCRadar’s stealer-log checks found no credential matches for many listed maritime organisations, and the data sets are small and application-scoped. The working hypothesis becomes an exposed shared application or vulnerability. No new technique is added; the access question for this batch is open, and the shared source is unconfirmed (CyberLogitec link unverified).
First mapping published. CyberXTron’s early table associates ten techniques with the group: T1078, T1059, T1021.001, T1547, T1074, T1486, T1490, T1489, T1027 and T1685. Procedures are not published at this stage. Note: “T1685 — Disable or Modify Tools” does not exist in MITRE ATT&CK; the behaviour matches T1562.001, confirmed later by SOCRadar, and T1685 is not carried as a technique on this page.
As published by SOCRadar from the operator’s exposed workspace (28 Oct 2026). Defanged. The full list on SOCRadar’s platform may be longer; this is the published set.
| Type | Indicator | Context |
|---|---|---|
| IPv4 | 185.157.161.148 | Agent box — kserver C2 — /x/i beacon; open file server on :8877 (AS42675, Sweden) |
| 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 |
| notice@n0n-cases.net | Extortion mail sender address | |
| Domain | n0n.to | Leak site (clearnet, inactive) |
| Domain | uberip.com | Actor mail domain |
| Onion | nongzecboljwv3yfndkggsybsglfrkffw7bvk2zemuteoxe6etpusnad.onion | Onion leak site (Tor) |
| Type | Indicator | Context |
|---|---|---|
| File / path | *.n0n, *.n0n.tmp, n0n_READ_ME.txt, C:\Users\Public\n0n\ | Encrypted files — ransom note — staging directory |
| Registry | HKLM\Software\Microsoft\Windows\CurrentVersion\Run : n0nNote | Ransom note at logon |
| Account | svc_upd (“Software Update”), svc_reg (“Registry Service”), svc<slug> | Hidden administrators |
| Scheduled task | [NAME]Sync*, [NAME]Audit, [NAME]Exfil, NetAssist, ChromeDbg* | SYSTEM tasks referencing C:\Users\Public\n0n\ |
| Firewall | Snort rule “n0n-lock” (sid 2100001) | Injected on compromised Cisco FMC appliances |
| SSH | authorized_keys entries ending root@3090-AI | Operator keys on FMC appliances and cloud clones |
| Type | Value | Context |
|---|---|---|
| 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 |
As published by Wiz, 9 Nov 2026. These indicators belong to the Qubole campaign; the link between that campaign and n0n is reported, not verified. Do not merge them with the SOCRadar indicators above.
| Type | Indicator | Context |
|---|---|---|
| AWS account | 805246085872 | Qubole’s platform account — hunt AssumeRole calls from non-AWS addresses |
| IAM principal | AIDA3W7DIFLYOLTK63SFS | Attacker-only in the environments Wiz examined |
| IAM principal | AIDAJYRRH5Y3A7AAJM2HW | Also carries legitimate Qubole traffic — only suspicious from non-AWS sources |
| IPv4 | 45.83.220.211, 185.209.199.73, 89.37.63.175 | Mullvad VPN exit nodes, shared — a match alone is not proof |
| IP | 57.129.54.120, 2001:41d0:70d:7800:: | OVH-hosted addresses used by the actor |
| Role session | quiam-* | Role session names, e.g. quiam-enum |
| User agent | TruffleHog | On GetCallerIdentity calls, paired with an indicator above |
As published by SOCRadar, 28 Oct 2026. A check of all 55 showed zero transactions as of 16 October 2026.
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
| Where to look | Marker |
|---|---|
| FortiGate / firewall logs | /remote/logincheck requests carrying Cookie: APBSID=n0n; bursts of failed logins with many distinct usernames from one source |
| Endpoint | Creation of C:\Users\Public\n0n\; *.n0n, *.n0n.tmp, n0n_READ_ME*; rapid renames to .n0n after MSSQL/W3SVC/MySQL termination and vssadmin shadow deletion |
| PowerShell | Net.WebClient.DownloadFile to HTTPS hex URLs with ServerCertificateValidationCallback={$true}, followed by Get-FileHash SHA1 checks |
| Persistence | SYSTEM scheduled tasks referencing C:\Users\Public\n0n\; Registry Run key n0nNote; hidden local admins with non-expiring passwords and UserList=0 |
| Browser | Chrome launched with --remote-debugging-port on cloned profiles; 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; authorized_keys entries ending root@3090-AI |
| Cloud | CloudTrail sequences of ListBuckets, sustained GetObject and DeleteObjects in 1,000-item batches; PutObject writing n0n_READ_ME.txt; ssm:SendCommand from new or inactive keys |
| Hosting platforms | New OnApp recipes or Zabbix scripts with “maintenance” names; backup-to-template conversions; VMs tagged mnt-* |
| Network | JSON POSTs to /x/i containing “ippay-web” with card data |
What the counts actually mean — and what they do not. Leak-site listings are the group’s own claims; a listing is not proof of a breach.
| Date | Claimed total | Basis |
|---|---|---|
| 22 Sep 2026 | 13 | Leak-site listings, 10 countries (CyberXTron) |
| 8 Oct 2026 | 21 | Leak-site listings, 17 countries (SOCRadar) |
| 12 Oct 2026 | 59 (21 + 38 maritime) | Single maritime batch across 18 countries (SOCRadar, Cydome) |
| 19 Oct 2026 | 93 | Leak-site listings; total demands $283.7M (SOCRadar) |
| Early Nov 2026 | 170+ | Claimed victims (Check Point Research) |
| 4 Nov 2026 | ~189 | Claimed since mid-September (OSec) |
| Mid-Nov 2026 | 195 | Claimed victims through mid-month (HackerFeeds, via trade press) |
| 16 Nov 2026 | 363 entries | Ransomtracker entries, of which 168 are numbered placeholders (Ransomnews) |
| — | Just over 20 | Real targets per the group’s own exposed server (Bitdefender, SOCRadar: 23 organisations in the workspace) |
Tracker totals disagree because they count different things — listings, named victims, placeholders, and workspace evidence. They are kept side by side here and never added together.
Early victimology is broad and opportunistic. The first 13 listings: financial services 23% (3), technology, retail & e-commerce and education 15% each (2), government & defence, healthcare, professional services and other 8% each (1). The workspace’s 23 organisations span 15 sectors. October concentrates in logistics and shipping (the 38-organisation maritime wave; South Korea accounts for 12 of the 38, including Busan terminal operators BNCT and Pusan Newport Company), and Bitdefender notes South Korean victims rose to 18 in October, 11 of them claimed by n0n and all in transportation. OSec’s assessment: the primary geographic focus is the US and South Korea, and the most-targeted sectors are logistics and shipping, technology, retail, healthcare, manufacturing and finance.
| Organisation | What n0n claims | Status |
|---|---|---|
| Visa | Cardmember programme records (Asia Pacific): names, emails, phone numbers, travel and dining bookings. No card numbers mentioned. | Reported — Data published 19 Nov; Cybernews examined the dump and found more than 6,000 premium customers affected, the data likely from a marketing platform for Visa’s premium dining programmes in Asia-Pacific. Visa confirmed an incident at a third-party provider in Asia Pacific; Visa’s systems not involved, no sign card numbers affected. n0n’s role unconfirmed. Ransomware.live, 18 Nov: most likely Qubole-sourced, not a direct breach (assessed, unconfirmed). |
| Mastercard | MEA cardmember programme records claimed. | Reported — Data published 19 Nov: about 2,000 entries in a 280KB file; Cybernews researchers say it looks like test data rather than genuine customer information, and its origin and authenticity are unverified. Ransomware.live, 18 Nov: most likely Qubole-sourced, not a direct breach (assessed, unconfirmed). |
| JPMorgan Chase | Email campaign records for a premium Reserve card programme claimed. | Reported — Data published 19 Nov: roughly 16,000 customer email addresses from the Reserve programme, phone numbers in some cases, account IDs and 124 employee email addresses. JPMorgan confirmed 19 Nov that third-party vendors were targeted and marketing data was accessed; services and sensitive financial data unaffected, and the attack is not connected to the 16 Nov outage. Ransomware.live, 18 Nov: most likely Qubole-sourced, not a direct breach (assessed, unconfirmed). |
| D.R. Johnson Financial | 245,667 distinct emails; 5,548 phone numbers; GoldMine archive of 89,740 files. | Reported — data published, per leak-site reporting. |
| Tailored Benefits | 434,983 emails claimed. | Unconfirmed |
| ARCA Financial; Opus Wealth Advisors; Pearl Planning; Terracotta Wealth; Wise Advisory Group | Canadian advisory and planning firms, named 16 Nov 2026. | Unconfirmed |
| Steeden Financial | 600,000+ emails claimed, including Social Insurance Numbers. | Unconfirmed |
| Scry Analytics | Leak of a US private banking client’s customer/signer records, plus an insurance group’s 7,904-record party master. | Unconfirmed — leak-site post only, no victim confirmation. |
| MM Insurance Agency | Listed as a claimed victim. | Unconfirmed |
| Higginbotham | 217,000+ documents published. | Reported — publication claimed; victim has not confirmed. |
| Two unnamed Canadian financial firms | Listed as claimed victims. | Unconfirmed |
| Transcom (PayPal’s support outsourcer) | 86.7 million support-agent records claimed from PayPal’s Citrix and authentication systems, plus an internal network map. | Unconfirmed |
| Expedia | Named 9 Nov 2026; claimed travel data. | Reported — Expedia confirmed a supplier breach on 12 Nov but did not confirm n0n’s figures or say n0n was responsible. Wiz and Sentra documented an October campaign abusing Qubole credentials; Expedia has not said its incident is related. |
| Ekwiik (Ideal Prepaid Botswana) | Prepaid utilities and airtime platform, Botswana; listed anonymously as “Company #216” since 13 Nov; no data volume or types disclosed. | Unconfirmed — leak-site listing only, no victim statement. |
Dediserve (cloud hosting): taken over through stolen admin browser credentials, with tenant backups extracted — the only Linux locker deployment in the workspace. 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.
The details that change how to think about this group.
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; gateway spending passed $7,216 by 22 October 2026. The evidence points to a human-directed operation supported by AI; no fully autonomous attack system was observed. The group’s own server also held the AI environment alongside 829 scripts and five encryptor sources.
Demands range from 0.5 BTC to 33 BTC per victim, and up to $28 million per organisation in the maritime wave (the reviewed subset ran $2.4–$14 million). 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 16 October 2026. One Canadian financial-services company carries the leak site’s “HONORED” label with a settlement date of 7 October 2026 — the only settlement marker found. The gap between the claimed scale and the observed payments is unresolved.
Each locker build uses a single hardcoded AES-256 key with no per-victim derivation, and one key is likely shared across two victims — so a decryptor for one could unlock the other. Bitdefender notes the same reuse. 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.
Pressure is not limited to the victim. n0n contacts victims’ clients, MSPs, unions and the media, and in one case emailed PayPal’s security team over a breach at its contact-centre provider. The maritime wave shows the same logic at scale: the Nam Dinh Vu Port posting states its archive would expose the cargo flows, booking relationships and operational records of every carrier that called at the terminal.
Listings carry a status and a countdown. “PENDING EXPOSURE” anonymises the victim as “Company #N” with only a sector and location, one to three days out. “NOT PAID, EXPOSED” names the victim with a publication date. “HONORED” marks a victim that settled before its deadline. Data is offered as a direct .zip archive and a .torrent, distributed from a VPN-bound seedbox. The maritime batch broke the earlier rhythm: named postings with a single shared publication time, seven days out. Data typically leaks about four days after public disclosure.
OSec notes a recent Fortinet brute-forcing campaign (dubbed “Fortibleed”) likely fuelled the group’s access. The group’s own credential-stuffing pipeline targeted 305 FortiGate SSL-VPN appliances through residential proxies. No Fortinet vulnerability is involved — the attacks use weak or stolen credentials.
Links reported by others, kept separate from the group’s own record until evidence ties them together.
CloudSEK’s “Caught in 4K: The Gentlemen Files” (5 Oct 2026) documents a Russian-speaking affiliate of The Gentlemen ransomware-as-a-service group, using the name Azazel, who used the group’s tooling, negotiation channels and ransom-note template but posted victims on his own leak site (LEAKNED) and kept the proceeds. His exposed servers showed 24+ victim directories and about 6 TB of staged data, and the first reported operational use of MCP’s exec_in_session as a command-and-control channel — a Penelope reverse-shell handler registered as a tool in an AI coding assistant, driven through a local MCP endpoint at 127.0.0.1:35367.
Aggregator detection content (Threadlinqs TL-2026-5634) associates this cluster with n0n at low confidence. The association is not adopted here: the two exposed workspaces are different servers with different toolsets (OpenCode and credential stuffing for n0n; Penelope and MCP for Azazel), SOCRadar’s n0n mapping contains no MCP technique, and Bitdefender treats the two as separate stories in the same debrief. If evidence later ties them together, it will be recorded in the TTP change log.
No definitive geographic or national attribution exists. Some threat-intelligence vendors list n0n as likely Russia-based, but the evidence behind that assessment has not been shown. Cydome notes evidence that the group is not related to Iran, China or North Korea, and a detail in the maritime analysis was read as a possible indication of Christian origin — neither is attribution. SOCRadar assesses with medium confidence that the exposed 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.
Drawn from SOCRadar, Fortra and Bitdefender. Prioritised for identity, remote access, cloud and backup owners.
| Area | Action |
|---|---|
| Remote access | Enforce phishing-resistant MFA on RDP, SSL-VPNs and admin portals; monitor /remote/logincheck for credential stuffing (hunt Cookie: APBSID=n0n); invalidate credentials found in infostealer logs; remove internet-exposed RDP. |
| Firewall management | Restrict FortiManager and FMC interfaces with IP allow-lists and MFA; require approvals for policy changes; flag REST token creation from unknown IPs; hunt the injected Snort rule n0n-lock (sid 2100001). |
| Cloud keys | Replace long-lived keys with short-lived roles; separate S3 list, read and delete permissions; enable Object Lock and versioning; alert on high-volume DeleteObjects calls. |
| Backups | Keep air-gapped or immutable backups and test restores — n0n purges shadow copies, wipes free space and deletes provider-hosted backups. |
| Credentials | Enforce strong password policy and regular rotation; hunt hidden admins (svc_*), UserList=0 accounts and root@3090-AI authorized_keys entries. |
| Detection | Search EDR, CloudTrail, WinRM/RDP and firewall logs for the indicators and behaviours in this page; consider restricting PowerShell, Python and other advanced scripting for users outside IT. |
| Extortion readiness | Prepare a data-breach and extortion-specific response plan, not just ransomware recovery — stolen data remains stolen even if systems are restored from backup. Ask suppliers first when a company appears on a leak site: much of n0n’s claimed data is the kind suppliers hold. |
Open questions, stated plainly rather than filled in.
Primary sources first. Leak-site content is not linked and no stolen data was accessed.
Page pattern: this profile follows the Tracking Offensive AI Tooling page. New findings go in the TTP change log at the top with a date; the matrix and procedure table update with them. First published 18 November 2026.