Threat reportPhishingTL-2026-2595
Large-Scale Azure-Hosted Tech Support Scam Campaign Targets Japan (13.38M Emails, 240K+ Relay IPs, 33K+ Disposable Azure Blob Storage Sites)
Between mid-December 2025 and May 2026, a sustained tech support scam campaign sent 13.38 million emails (94% to .jp addresses) via 240,000+ globally distributed sending/relaying IP addresses, driving victims to 33,000+ disposable fake security-alert pages hosted on Microsoft Azure Blob Storage static website hosting.
Key facts
| Identifiers | TL-2026-2595 |
|---|---|
| Severity | HIGH |
| Affected products | Microsoft Azure Blob Storage (static website hosting); MikroTik RouterOS-based devices with internet-exposed management/bandwidth-test services; Multiple Individual and organizational email users at .jp domains (Japan) |
| Malware | AnyDesk, TeamViewer |
| Tools | AnyDesk, LogMeIn, RustDesk, ScreenConnect, TeamViewer, UltraViewer |
| MITRE ATT&CK techniques | T1027 Obfuscated Files or Information, T1036.005 Match Legitimate Resource Name or Location, T1204.001 Malicious Link, T1204.002 Malicious File, T1219 Remote Access Tools, T1566.002 Spearphishing Link, T1566.004 Spearphishing Voice, T1583.006 Web Services, T1584.005 Botnet, T1589.002 Email Addresses, T1598.003 Spearphishing Link, T1657 Financial Theft and 1 more in the attack flow |
| Motivation | FINANCIAL |
| Target sectors | consumer, financial services impersonated, ecommerce, government tax authority impersonated, cross-sector enterprise workplace-themed phase |
| Target regions | japan, Asia-Pacific |
| Detection rules | 9 (1 shown free) |
| Sources | 6 cited sources |
Detection coverage
As of 2026-09-21, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-2595 across Splunk SPL, Microsoft KQL and Sigma, covering 20 indicator(s) of compromise. The whole corpus is readable without an account. Every threat comes with one complete detection in Splunk SPL, Microsoft KQL and Sigma, free for everyone; a free account adds daily TL Creds to unlock the rest. Paid plans add a larger daily allowance, correlation and the MCP server. Threadlinqs MCP server · View plans.
Free detection: Self-Spoofed Sender Phishing Email Linking to Azure Blob Storage Fake Alert Page
Detects inbound email where the sender address is spoofed to match the recipient's own address (used in >90% of this campaign's lures) and the message or a linked URL points to Azure Blob Storage static website hosting (*.blob.core.windows.net / *.z##.web.core.windows.net), the platform used for 33,000+ disposable tech-support-scam landing pages targeting .jp addresses.
- Languages: Splunk SPL, Microsoft KQL, Sigma
- ATT&CK: T1566.002, T1656, T1589.002, T1583.006
- Data sources: email_gateway_logs, microsoft_defender_office365_EmailEvents, microsoft_defender_office365_UrlClickEvents
- Severity: high
- Confidence: high
Splunk SPL
index=email sourcetype="email:security" (mail_from=mail_to OR envelope_from=envelope_to)
| eval spoofed_self_sender=if(lower(mail_from)=lower(mail_to) OR lower(envelope_from)=lower(envelope_to),1,0)
| where spoofed_self_sender=1
| eval blob_link=if(match(url,"(?i)(\.blob\.core\.windows\.net|\.z\d+\.web\.core\.windows\.net)"),1,0)
| where blob_link=1 OR match(body,"(?i)(\.blob\.core\.windows\.net|\.z\d+\.web\.core\.windows\.net)")
| eval jp_target=if(match(mail_to,"(?i)\.jp$"),1,0)
| table _time, mail_from, mail_to, subject, url, src_ip, jp_target
| stats count min(_time) as first_seen max(_time) as last_seen values(url) as urls by mail_from, mail_to, subject
| where count > 0Microsoft KQL
let phishHosts = dynamic(["blob.core.windows.net", "web.core.windows.net"]);
EmailEvents
| where Timestamp > ago(24h)
| where SenderFromAddress =~ RecipientEmailAddress
| join kind=inner (
UrlClickEvents
| where Timestamp > ago(24h)
| where Url has_any (phishHosts)
) on NetworkMessageId
| project Timestamp, SenderFromAddress, RecipientEmailAddress, Subject, Url, ActionType, NetworkMessageId
| where RecipientEmailAddress endswith ".jp" or isnotempty(RecipientEmailAddress)Sigma
title: Self-Spoofed Sender Phishing Email to Azure Blob Storage Landing Page
id: 7a1f2c3e-4b5d-4e6f-9a1b-2c3d4e5f6a71
status: experimental
description: Detects phishing email where sender is spoofed to the recipient's own address, linking to an Azure Blob Storage static-website phishing page, consistent with the TL-2026-2595 Japan tech-support-scam campaign.
references:
- https://www.trendmicro.com/en_us/research/26/g/tech-support-scams-targeting-japan.html
author: Threadlinqs Team
date: 2026-09-21
tags:
- attack.reconnaissance
- attack.t1589.002
- attack.initial-access
- attack.t1566.002
- attack.defense-evasion
- attack.t1656
logsource:
category: mail
product: email
detection:
selection_spoof:
sender_address|fieldref: recipient_address
selection_blob:
target_url|contains:
- '.blob.core.windows.net'
- '.web.core.windows.net'
condition: selection_spoof and selection_blob
falsepositives:
- Internal automated notification systems that legitimately send-to-self via a shared mailbox
- Legitimate Azure Blob Storage static-site links from a known internal SaaS integration
level: highKnown false positives:
- Legitimate 'send-to-self' automated workflows (ticketing confirmations, calendar invites) sharing a mailbox address
- Internal SaaS platforms that legitimately host static content on the organization's own Azure Blob Storage account
Tuning notes: Scope RecipientEmailAddress to the organization's .jp-facing mail domains first to cut noise, then expand; add an allow-list for the org's own verified Azure Storage account hostnames to avoid false-positiving on legitimate internal static-site links.
- User Click-Through to Azure Blob Storage Static-Website Phishing Page · unlock with TL Creds
- Remote-Access Tool Installer Executed with Browser/Explorer Parent Process · unlock with TL Creds
- RAT C2 Network Connection Correlated with Prior Scam Callback (Active Session Confirmation) · unlock with TL Creds
- Sender-Recipient Address Match on Inbound Mail (Self-Spoofing Phishing Indicator) · unlock with TL Creds
- Consumer Remote-Access Tool Installed Outside Software Deployment Baseline · unlock with TL Creds
- Network Connection to Remote-Access Tool Relay/C2 Infrastructure · unlock with TL Creds
- Outbound Call to Known Tech-Support-Scam Callback Number (010 + NANP Pattern) · unlock with TL Creds
- Remote-Access Tool C2 Session Followed by Banking Portal Navigation (Financial Theft Indicator) · unlock with TL Creds
Indicators of compromise
As of 2026-09-21, Threadlinqs Intelligence tracks 20 indicators of compromise for TL-2026-2595.
Technical analysis
Victims who call the displayed North American (+1) support numbers are talked into installing legitimate remote-access software (AnyDesk, TeamViewer, RustDesk, ScreenConnect, UltraViewer, LogMeIn), leading to credential theft and fraudulent banking transactions.
Trend Micro tracked a 165-day tech support scam operation that sent 13.38 million emails at an average of ~81,000/day, peaking at 4.45 million emails (~160,000/day) in February 2026 and settling to a sustained ~30,000/day by May 2026. Ninety-four percent of targeted addresses used Japan's .jp top-level domain, and delivery timing was tuned to Japan Standard Time business hours (09:00-21:00 JST).
Delivery infrastructure comprised more than 240,000 sending/relaying IP addresses distributed globally (Brazil contributed the largest email volume; China the most unique source IPs). A subset of this infrastructure fingerprinted as legitimate, internet-exposed network devices rather than dedicated attacker infrastructure: 3,231 relaying IPs ran MikroTik services, of which 2,657 exposed MikroTik's TCP/2000 (bandwidth-test) service to the internet, alongside 284 exposing OpenSSH, 184 exposing nginx, and 144 exposing Apache. This fingerprint is consistent with hijacked/unmanaged IoT and small-office router equipment being recruited as spam relays, a pattern independently documented in unrelated 2025 MikroTik-botnet malspam reporting.
Landing infrastructure used Microsoft Azure Blob Storage's static website hosting feature to serve more than 33,000 disposable fake security-alert pages; roughly 90% were used for only a single day before rotation, and page content was encrypted to evade content-based scanners. Abuse of Azure Blob Storage's trusted, Microsoft-issued-TLS static hosting for phishing/scam landing pages mirrors broader 2025-2026 industry reporting on cloud-native phishing infrastructure abuse.
The campaign's lures evolved in two phases. From mid-December 2025 through March 2026, lures were generic: fake security warnings, adult-content pretexts, and impersonation of e-commerce platforms, financial institutions, and tax agencies. From April 2026 onward, the operators shifted to workplace-themed social engineering aimed at individuals inside organizations, with subject lines such as "[Urgent] Internal network security audit: Request to verify suspicious device activity and logs," "[Confidential] Advance release of the H2 FY2026 performance evaluations and promotion candidate list," "[Important / All employees] Confirmation of H2 FY2026 salary revisions and evaluation feedback," "[Employee benefits] Notice: Digital Amazon gift cards for all employees," "[Important notice] Changes to commuting expense reimbursement rules," and "[Urgent] Request for confirmation regarding flat-rate tax cut and refund procedures." More than 90% of sender addresses were spoofed to match the recipient's own address, a technique intended to defeat casual sender-legitimacy checks and increase open rates.
The monetization chain is classic callback phishing / Telephone-Oriented Attack Delivery (TOAD): the email or landing page displays a support phone number; the victim calls it; a live operator instructs the victim to install a remote-access tool (AnyDesk, TeamViewer, RustDesk, ScreenConnect, UltraViewer, or LogMeIn); the operator then uses that access to harvest credentials, navigate the victim into their online banking portal, and extract fraudulent fees or transfers. Trend Micro identified 11-12 distinct callback phone numbers reused across 4,721 of the fake alert sites (February-May 2026 data, current as of June 12, 2026); the numbers use Japan's "010" international access prefix followed by North American (+1) numbers, indicating a North America-based call operation servicing Japanese victims.
Japan's National Police Agency's 2025 fraud statistics show "support-pretext" billing fraud (the category corresponding to tech support scams) at 1,048 reported cases (down 31.2% year-over-year) but 1.49 billion yen in losses (up 48.1% year-over-year) -- meaning far fewer victims are being defrauded, but for roughly double the average amount per case, consistent with the campaign's April 2026 pivot toward higher-value enterprise/workplace targets.
No threat-actor attribution is offered in available reporting. This campaign is a distinct operation from, but part of the same broader tech-support-scam ecosystem as, Barracuda's May 2026 "CypherLoc" browser-locking scareware kit report; the two should not be conflated as the same infrastructure or toolset absent further evidence.
MITRE ATT&CK techniques by tactic:
Defense Evasion
T1027 Obfuscated Files or Information; T1036.005 Match Legitimate Resource Name or Location; T1684.001 Impersonation
Execution
T1204.001 Malicious Link; T1204.002 Malicious File
Command and Control
Initial Access
T1566.002 Spearphishing Link; T1566.004 Spearphishing Voice
Resource Development
T1583.006 Web Services; T1584.005 Botnet
Reconnaissance
T1589.002 Email Addresses; T1598.003 Spearphishing Link
Impact
Remediation
Immediate actions
- Enforce SPF, DKIM, and DMARC (with a reject/quarantine policy) on all inbound mail flows to catch senders spoofed to match the recipient's own address or internal domains
- Block or restrict installation of consumer/SMB remote-access tools (AnyDesk, TeamViewer, RustDesk, ScreenConnect, UltraViewer, LogMeIn) on non-IT-managed endpoints via application allow-listing
- Alert on outbound international calls dialed via Japan's '010' access prefix to North American (+1) numbers occurring shortly after a user clicks a security-alert email link
- Brief HR, payroll, and IT helpdesk teams on the workplace-themed subject-line lures (performance review, salary revision, security audit, gift card, expense-policy, tax-refund pretexts) currently in use
Workarounds
- Require employees to verify unsolicited 'urgent' HR/IT/security emails through a known internal channel (ticketing system, verified intranet contact) before calling any phone number contained in the email or a linked page
- Restrict or monitor outbound international voice calling for employees where feasible, particularly to the North American numbers identified as IOCs for this campaign
Longer-term hardening
- Deploy multi-layer email security with URL and page-content inspection capable of following links to Azure Blob Storage static-website hosts (*.blob.core.windows.net / *.z##.web.core.windows.net) rather than trusting the domain by reputation alone
- Inventory and remediate internet-exposed MikroTik/IoT and small-office router management interfaces (e.g., TCP/2000 bandwidth-test service, exposed SSH/HTTP admin panels) that could be recruited as spam-relay infrastructure
- Run recurring, scenario-based security-awareness training emphasizing urgency-based social engineering and callback-phishing (TOAD) tactics, not just malicious-link recognition
- Establish a verified callback-number registry for internal IT/HR/security communications so employees can distinguish legitimate outreach from spoofed workplace-themed lures
Affected and fixed versions
- Microsoft — Azure Blob Storage (static website hosting)
Vulnerable versions: N/A -- abused as trusted, Microsoft-TLS-backed hosting infrastructure for disposable phishing/scam landing pages, not a software vulnerability
Fixed in: N/A - MikroTik — RouterOS-based devices with internet-exposed management/bandwidth-test services
Vulnerable versions: Devices with internet-accessible TCP/2000 (bandwidth-test), SSH, or HTTP admin services, consistent with unmanaged/hijacked units recruited as spam relays
Fixed in: N/A -- restrict management-plane exposure to the internet - Multiple — Individual and organizational email users at .jp domains (Japan)
Vulnerable versions: N/A -- social-engineering campaign targeting end users, not a software product
Fixed in: N/A
Timeline
- Campaign begins sending fake security-alert emails to Japanese (.jp) addresses, linking to disposable landing pages hosted on Microsoft Azure Blob Storage static website hosting.
- Campaign reaches its peak monthly volume of 4.45 million emails (~160,000/day) in February 2026.
- Japan Times reports 2025 fraud figures showing specialized/support-pretext fraud losses at record highs nationally, providing corroborating context for the campaign's financial-fraud outcomes.
- Lures pivot from generic security/adult-content/financial-institution/tax-agency impersonation to workplace-themed subject lines targeting employees inside organizations (performance reviews, salary revisions, internal security audits, gift-card notices, expense-policy changes, tax-refund pretexts).
- Campaign volume declines from its February peak to a sustained ~30,000 emails/day while remaining active.
- Barracuda Networks separately publishes a report on 'CypherLoc,' a browser-locking scareware kit that also drives victims to fraudulent tech-support calls, illustrating the broader tech-support-scam ecosystem active during the same period (distinct infrastructure/toolset from this campaign).
- Trend Micro's tracked set of reused scam callback phone numbers (11-12 numbers across 4,721 fake alert sites, February-May 2026) is current as of this date.
- Trend Micro publishes 'Tech Support Scams Targeting Japan,' detailing the campaign's infrastructure, lure evolution, and scale.
Sources cited
- 13M+ Emails Sent in Tech Support Scam Targeting Users, Organizations in Japan Trend Micro · Research · 2026-07-23
- Inside the attack chain: Threat activity targeting Azure Blob Storage Microsoft · Vendor report · 2025-10-20
- Threat Spotlight: CypherLoc, an advanced browser-locking scareware targeting millions blog.barracuda.com · Research · 2026-05-20
- 13,000 MikroTik Routers Hijacked by Botnet for Malspam and Cyberattacks The Hacker News · News · 2025-01-21
- Fraud soars to record high in Japan as scam tactics grow more sophisticated japantimes.co.jp · News · 2026-02-12
- Fake Police Scams Double in Japan as Financial Losses from Fraud Soar nippon.com · News · 2026-06-17
Previous and next in phishing
Drafted by the Threadlinqs research pipeline from 6 sources; published ; last updated —.