# Credential-harvesting phishing page collecting victim MX records via DNS-over-HTTPS, hosted on Google Cloud Storage

> A password-expiry phishing lure leads to a credential-harvesting page hosted on storage.googleapis.com. Besides the password, the kit collects public IP, geolocation, browser info, language and timestamp, plus the MX record of the victim's email domain via a DNS-over-HTTPS lookup to dns.google, and exfiltrates everything through the Telegram Bot API. SANS ISC reports the MX data was not used to customize the page.

- **Published:** 2026-11-17T11:14:42Z
- **Last reviewed:** 2026-11-17T11:14:42Z
- **Canonical:** https://intel.threadlinqs.com/threat/TL-2026-6238
- **Card image:** https://intel.threadlinqs.com/og/threat/TL-2026-6238.png
- **ID:** TL-2026-6238
- **Severity:** MEDIUM
- **Category:** PHISHING
- **Status:** ACTIVE
- **Detections:** 9 (1 shown free below) · **IOCs:** 10 (3 shown free below)

## Description

SANS ISC handler Jan Kopriva analyzed a phishing campaign received over several weeks. The email is a generic 'your password will expire today' lure. The link points to a page on Google Cloud Storage (storage.googleapis.com/alx2/03/1a/we.html) with the recipient's email address in the URL fragment. The page renders a generic login form personalized with the domain of the recipient's address and uses a screenshot of the target organization's website as background, fetched via the thum.io screenshot service. The author notes this background failed for isc.sans.edu but worked convincingly for another domain.

The page code is hidden behind two layers of URL encoding and additional JavaScript obfuscation. After deobfuscation, the kit supports multiple languages and, on submission, collects the password, the visitor's public IP address, geographic location, browser information, language and a timestamp. It also performs a DNS-over-HTTPS lookup for the MX record of the victim's email domain against Google's public resolver (dns.google/resolve, type MX) and adds an 'MX Record' field to the stolen-data message. The data is exfiltrated to the attacker through the Telegram Bot API. The diary does not disclose the bot token or chat ID.

The MX lookup does not appear to customize the page. The author says it could help an attacker identify the target's email infrastructure or prioritize credentials, but this cannot be confirmed. Related prior reporting is Infoblox's March 2025 research on Morphing Meerkat, a phishing-as-a-service operation that used DoH MX queries to select among 114 brand templates, relied on open redirects and compromised sites, and could exfiltrate via a Telegram bot webhook. No link between this campaign and Morphing Meerkat is stated in the source, and no actor, malware family, IP address or hash is reported. Severity and category are analyst-assigned. No confirmed victims are reported.

Defender relevance: the kit abuses trusted cloud and utility services (Google Cloud Storage, dns.google, thum.io, Telegram), so domain reputation alone will not catch it. Useful hunting pivots are browser or proxy traffic to dns.google/resolve with type=MX, requests to api.telegram.org from non-Telegram clients, and storage.googleapis.com URLs carrying an email address in the fragment.

## MITRE ATT&CK

- T1598 Phishing for Information
- T1590 Gather Victim Network Information
- T1589 Gather Victim Identity Information
- T1583 Acquire Infrastructure
- T1566 Phishing
- T1204 User Execution
- T1027 Obfuscated Files or Information
- T1140 Deobfuscate/Decode Files or Information
- T1071 Application Layer Protocol
- T1567 Exfiltration Over Web Service

## Free detection (Splunk SPL, Microsoft KQL, Sigma)

**Password-Expiry Phishing Page on Google Cloud Storage (storage.googleapis.com/alx2/03/1a/we.html) — Exploitation** (`TL-2026-6238-SPL-001`)

Behaviour-first detection of the TL-2026-6238 credential-harvesting landing page: proxy requests to the abused storage.googleapis.com host for the reported /alx2/03/1a/we.html object or /alx2/ path, plus a behavioural branch for Google Cloud Storage HTML pages opened from webmail clients. Block by URL path, not by domain, because storage.googleapis.com is legitimate. Maps to Phishing (T1566), User Execution (T1204) and Acquire Infrastructure: Web Services (T1583).

- **Languages:** Splunk SPL, Microsoft KQL, Sigma
- **ATT&CK:** T1566, T1204, T1583
- **Data sources:** proxy: index=proxy sourcetype pan:url/zscalernss-web/bluecoat:proxysg:access:syslog/squid:access, CommonSecurityLog (CEF web proxy), Sigma logsource category proxy
- **Severity:** high
- **Confidence:** high

### Splunk SPL

```spl
index=proxy sourcetype IN ("pan:url","zscalernss-web","bluecoat:proxysg:access:syslog","squid:access") earliest=-24h
    (url="*storage.googleapis.com/alx2/*" OR url="*/alx2/03/1a/we.html*" OR (url="*storage.googleapis.com/*.html*" http_referrer IN ("*mail.google.com*","*outlook.office.com*","*outlook.live.com*","*mail.yahoo.com*")))
| eval match_type=case(like(url,"%/alx2/03/1a/we.html%"),"exact_ioc_we.html", like(url,"%/alx2/%"),"alx2_path_pattern", true(),"gcs_html_from_webmail")
| stats count as hits min(_time) as first_seen max(_time) as last_seen values(url) as urls values(match_type) as match_types values(http_referrer) as referrers by src user
| convert ctime(first_seen) ctime(last_seen)
| sort - hits
```

### Microsoft KQL

```kql
CommonSecurityLog
| where TimeGenerated > ago(24h)
| where RequestURL has "storage.googleapis.com"
| extend MatchType = case(
    RequestURL contains "/alx2/03/1a/we.html", "exact_ioc_we.html",
    RequestURL contains "/alx2/", "alx2_path_pattern",
    RequestURL matches regex @"(?i)\.html#[^/\s]*@[^/\s]+", "email_in_fragment",
    "")
| where isnotempty(MatchType)
| project TimeGenerated, DeviceVendor, DeviceProduct, SourceIP, SourceUserName, RequestURL, RequestMethod, MatchType
```

### Sigma

```yaml
title: Password-Expiry Phishing Page on Google Cloud Storage (alx2/we.html)
id: 6f1c2a52-8d0e-4c1b-9a63-2b7d5e0f1a01
status: experimental
description: Detects web proxy requests to the Google Cloud Storage object path /alx2/03/1a/we.html (or the /alx2/ bucket path) serving a credential-harvesting password-expiry phishing page (TL-2026-6238).
references:
    - https://isc.sans.edu/diary/rss/33526
    - https://www.infoblox.com/blog/threat-intelligence/a-phishing-tale-of-doh-and-dns-mx-abuse/
author: Threadlinqs Team
date: 2026-11-17
tags:
    - attack.initial-access
    - attack.t1566
    - attack.execution
    - attack.t1204
    - attack.resource-development
    - attack.t1583
logsource:
    category: proxy
detection:
    selection_host:
        cs-host: 'storage.googleapis.com'
    selection_exact:
        c-uri|contains: '/alx2/03/1a/we.html'
    selection_pattern:
        c-uri|contains: '/alx2/'
    condition: selection_host and 1 of selection_*
falsepositives:
    - Legitimate Google Cloud Storage buckets that happen to use an /alx2/ prefix
level: high

```

**Known false positives:**

- Legitimate Google Cloud Storage buckets that use an /alx2/ object prefix
- Users opening legitimate HTML documents hosted on storage.googleapis.com from webmail

**Tuning notes:** Requires TLS inspection or URL-category logging so the full path is visible. Add the CIM sourcetypes actually used in your proxy tier. The URL fragment (recipient email) is never sent to the server, so rely on the path match. Drop the webmail-referrer branch if it is noisy; allowlist known internal GCS buckets by exact path.

### 8 more detection rules (Splunk SPL, Microsoft KQL, Sigma), unlock with TL Creds

- Email Click-Through to Google Cloud Storage Phishing Page (alx2 / Email in URL Fragment) (locked: unlock with TL Creds)
- Browser DoH Query to [indicator] Shortly After Google Cloud Storage Page Access — MX Recon (locked: unlock with TL Creds)
- Browser Connection to Telegram Bot API ([indicator]) — Credential Exfiltration (locked: unlock with TL Creds)
- [indicator] Screenshot Background Requested by a Google Cloud Storage Page — Phishing Page Render (locked: unlock with TL Creds)
- DoH MX Lookup Initiated from a [indicator] Page — Phishing Kit Recon (locked: unlock with TL Creds)
- Telegram Bot API Message with 'MX Record' Field — Phishing Kit Data Exfiltration (locked: unlock with TL Creds)
- Browser-Side DoH MX Lookup via [indicator]/resolve — Victim Email Infrastructure Recon (locked: unlock with TL Creds)
- Telegram Bot API Exfiltration ([indicator]) from Non-Telegram Client — C2/Exfil (locked: unlock with TL Creds)

Sign up free to unlock the detection rules: https://threadlinqs.com/?modal=signup

## Sample indicators (3 of 10)

Defanged, TLP:CLEAR.

- Domain: `storage[.]googleapis[.]com`
- Domain: `dns[.]google`
- Domain: `thum[.]io`

Full set of 10 indicators unlocks with TL Creds: https://threadlinqs.com/tl-creds

## Sources

- [SANS ISC Diary: A phishing page interested in MX records](https://isc.sans.edu/diary/rss/33526)
- [Infoblox: A phishing tale of DoH and DNS MX abuse (Morphing Meerkat)](https://www.infoblox.com/blog/threat-intelligence/a-phishing-tale-of-doh-and-dns-mx-abuse/)
- [The Hacker News: New Morphing Meerkat phishing kit](https://thehackernews.com/2025/03/new-morphing-meerkat-phishing-kit.html)
- [Security Affairs: Morphing Meerkat phishing kits exploit DNS MX records](https://securityaffairs.com/?p=176029)
- [SecurityWeek: Morphing Meerkat phishing kits target over 100 brands](https://www.securityweek.com/morphing-meerkat-phishing-kits-target-over-100-brands/)
- [ITWire: Morphing Meerkat spoofs over 100 brands and steals credentials](https://itwire.com/business-it-news/security/meerkat-mayhem-phishing-as-a-service-unleashed)

## Full data

The other 8 detection rules (Splunk SPL / Microsoft KQL / Sigma) unlock with TL Creds; a free account includes daily creds: https://threadlinqs.com/?modal=signup
Indicator values unlock with TL Creds, at a lower price on Red and Purple; AI agents can query rules and IOCs through the Threadlinqs MCP server (Purple): https://intel.threadlinqs.com/mcp

Canonical: https://intel.threadlinqs.com/threat/TL-2026-6238
