Threat reportPhishingTL-2026-6051
Threat Actors Use Google Ads To Target Ledger Users (Seed-Phrase Phishing via Fake Ledger Site)
In August 2026, Zscaler ThreatLabz analyzed a phishing campaign that used fraudulent sponsored Google ads, placed under a long-standing verified advertiser profile (possibly a compromised account), to target Ledger hardware wallet users. Victims were redirected through Google Cloud Storage and rotating Vercel domains to a Google Sites page embedding a fake Ledger site that harvests secret recovery phrases, enabling wallet theft without the physical device. First disclosed .
Key facts
| Identifiers | TL-2026-6051 |
|---|---|
| Severity | HIGH |
| First disclosed | |
| Affected products | Ledger Ledger hardware wallet users (Ledger Wallet / Ledger Live impersonation) |
| Malware | HTML.Phish.Ledger |
| Tools | hCaptcha (invisible mode) |
| MITRE ATT&CK techniques | T1204.001 Malicious Link, T1497.002 User Activity Based Checks, T1567 Exfiltration Over Web Service, T1583.006 Web Services, T1583.008 Malvertising, T1657 Financial Theft, T1684.001 Impersonation |
| Motivation | FINANCIAL |
| Target sectors | cryptocurrency, finance, individual consumers |
| Target regions | North America, Europe, Asia |
| Detection rules | 9 (1 shown free) |
| Sources | 7 cited sources |
Detection coverage
As of 2026-11-14, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-6051 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: TL-2026-6051 Ledger Google Ads Phish — Redirect Chain via GCS Bucket, Google Sites and Vercel (Exploitation)
Detects web-proxy traffic to the known infrastructure of the Ledger seed-phrase phishing campaign (Zscaler HTML.Phish.Ledger): the GCS landing buckets (apf-leg-ad-*, ledg-leg1-*), the Google Sites iframe wrapper pages (start-ledger-wallet, apps-ledger-wallet, download-ledger-wallet-pc, or any sites.google.com/view/*ledger* page) and the published rotating *.vercel.app hosts. Each client is tagged with the chain stages it reached (GCS landing, Google Sites iframe, Vercel phishing page); a client reaching two or more stages has followed the sponsored-ad redirect chain and is a likely victim of the fake Ledger page that harvests the BIP-39 recovery phrase.
- Languages: Splunk SPL, Microsoft KQL, Sigma
- ATT&CK: T1583.008, T1583.006, T1204.001, T1656
- Data sources: proxy:zscalernss-web, proxy:bluecoat:proxysg:access:syslog, proxy:squid:access, CommonSecurityLog
- Severity: high
- Confidence: high
Splunk SPL
index=proxy sourcetype IN ("zscalernss-web","bluecoat:proxysg:access:syslog","squid:access") earliest=-24h
| eval full_url=lower(coalesce(url, cs_uri, request_url, uri)), client=coalesce(user, cs_username, src_user, src_ip)
| where match(full_url, "storage[.]googleapis[.]com/(apf-leg-ad|ledg-leg1)-[0-9]+") OR match(full_url, "sites[.]google[.]com/view/[a-z0-9-]*ledger[a-z0-9-]*") OR match(full_url, "(soyyoo-cwpc5n0e|rpc-gbz5|whyavc-qwmv6stx|router-wdoi|node-f1ey)[.]vercel[.]app")
| eval stage=case(match(full_url, "storage[.]googleapis[.]com"), "1_gcs_landing", match(full_url, "sites[.]google[.]com"), "2_google_sites_iframe", true(), "3_vercel_phish")
| stats count min(_time) as first_seen max(_time) as last_seen values(stage) as stages values(full_url) as urls dc(stage) as stage_count by client
| eval first_seen=strftime(first_seen, "%F %T"), last_seen=strftime(last_seen, "%F %T")
| sort - stage_count - countMicrosoft KQL
CommonSecurityLog
| where TimeGenerated > ago(24h)
| extend FullUrl = tolower(RequestURL)
| where FullUrl matches regex 'storage[.]googleapis[.]com/(apf-leg-ad|ledg-leg1)-[0-9]+'
or FullUrl matches regex 'sites[.]google[.]com/view/[a-z0-9-]*ledger[a-z0-9-]*'
or FullUrl matches regex '(soyyoo-cwpc5n0e|rpc-gbz5|whyavc-qwmv6stx|router-wdoi|node-f1ey)[.]vercel[.]app'
| extend Stage = case(FullUrl contains 'storage.googleapis.com', '1_gcs_landing', FullUrl contains 'sites.google.com', '2_google_sites_iframe', '3_vercel_phish')
| summarize Hits=count(), FirstSeen=min(TimeGenerated), LastSeen=max(TimeGenerated), Stages=make_set(Stage), Urls=make_set(FullUrl, 10) by SourceIP, SourceUserName
| extend StageCount = array_length(Stages)
| order by StageCount desc, Hits descSigma
title: Ledger Seed-Phrase Phishing Redirect Chain via GCS, Google Sites and Vercel (TL-2026-6051)
id: 5b1e7c2a-3d4f-4a68-9c1b-6f0a2d8e4b11
status: experimental
description: Detects proxy requests to the GCS landing buckets, Google Sites iframe wrapper pages and rotating Vercel hosts used by the Google Ads Ledger recovery-phrase phishing campaign.
references:
- https://www.zscaler.com/blogs/security-research/threat-actors-use-google-ads-target-ledger-users
- https://github.com/ThreatLabz/iocs/blob/main/googleads_ledger_attack/IOCs.txt
author: Threadlinqs Team
date: 2026/11/14
tags:
- attack.resource-development
- attack.t1583.008
- attack.t1583.006
- attack.execution
- attack.t1204.001
- attack.defense-evasion
- attack.t1656
logsource:
category: proxy
detection:
selection_gcs:
cs-host: storage.googleapis.com
c-uri|startswith:
- '/apf-leg-ad-'
- '/ledg-leg1-'
selection_sites:
cs-host: sites.google.com
c-uri|re: '^/view/[a-z0-9-]*ledger'
selection_vercel:
cs-host:
- soyyoo-cwpc5n0e.vercel.app
- rpc-gbz5.vercel.app
- whyavc-qwmv6stx.vercel.app
- router-wdoi.vercel.app
- node-f1ey.vercel.app
condition: 1 of selection_*
falsepositives:
- Security teams or researchers analysing the campaign
- Legitimate community pages on Google Sites that mention Ledger in the URL
level: highKnown false positives:
- Security analysts or sandbox/URL-detonation systems visiting the published IOCs
- Legitimate community Google Sites pages that happen to contain 'ledger' in the URL slug (for example accounting-ledger sites)
Tuning notes: Adjust the index/sourcetype list and field aliases (url, cs_uri, src_ip, user) to your proxy. Add new GCS buckets, Google Sites slugs and *.vercel.app hosts from the ThreatLabz googleads_ledger_attack IOC repository (several hundred rotating hosts) via a lookup. Prioritise clients with stage_count >= 2. Exclude security-team scanner source IPs. The Google Sites regex is broad on purpose; tighten to the three known slugs if accounting-ledger sites cause noise.
- TL-2026-6051 Ledger Malvertising — Google Ad Click Landing on GCS Bucket (leg*-NNNNN Pattern) · unlock with TL Creds
- TL-2026-6051 Fake Ledger Google Sites Visit Followed by Repeated POSTs to Vercel (Invalid-Seed Resubmission) · unlock with TL Creds
- TL-2026-6051 Ledger Phish Domain Rotation — Client Resolving Multiple Randomly Named Vercel Hosts (DNS) · unlock with TL Creds
- TL-2026-6051 Fake Ledger Wallet Page Hosted on Google Sites (iframe Wrapper) · unlock with TL Creds
- TL-2026-6051 Fake Ledger Page — Cloudflare Web Analytics Beacon and hCaptcha Loaded From Random Vercel Page (Bot-Evasion Stack) · unlock with TL Creds
- TL-2026-6051 DNS Lookup of Published Ledger Phishing Vercel Domains and Community-Observed Malvertising Siblings · unlock with TL Creds
- TL-2026-6051 Fake Ledger Page — BIP-39 Wordlist Fetch Followed by POST to Vercel (Recovery-Phrase Submission) · unlock with TL Creds
- TL-2026-6051 Ledger Phish Infrastructure Rotation — GCS Bucket Redirecting to Rotating Vercel Hosts · unlock with TL Creds
Indicators of compromise
As of 2026-11-14, Threadlinqs Intelligence tracks 20 indicators of compromise for TL-2026-6051.
Technical analysis
Zscaler ThreatLabz analyzed a malvertising campaign in August 2026 (blog published 2026-09-25) in which fraudulent Google sponsored search ads impersonating Ledger hardware wallet support were shown to users in the United States, Europe and parts of Asia. The ads ran under a long-standing, Google-verified advertiser account with no observed prior malicious ads and a verified identity located in Germany; ThreatLabz suggests the account may have been compromised. The ad displayed 'google.com' and '10L+ visits in the past month', a figure that appears to refer to google.com rather than the phishing destination.
Attack chain: the victim clicks the sponsored ad and lands on a Google Cloud Storage (GCS) URL. The GCS page uses a JavaScript redirect to a Vercel-hosted page on a Vercel domain that changed roughly every 15-20 minutes, which defeats reputation-based blocking. The Vercel page then redirects to a Google Sites page that embeds the phishing page in an iframe. The fake Ledger interface offers download options for Windows, macOS, Linux and mobile, asks the user to select a device type, shows 'Connecting your Ledger' and 'Initializing Firmware Update' messages, and asks the user to confirm device ownership before prompting for the secret recovery phrase (SRP).
Credential harvesting and evasion: the page loads the 2,048-word BIP-39 English wordlist (api/bip39-english.txt) to give autocomplete in the recovery phrase fields, and initializes hCaptcha in invisible mode on submission. It collects device metadata and monitors keypresses, touches and mouse movements, sending them to a Vercel endpoint; ThreatLabz notes this could help distinguish real visitors from automated analysis tools. A Cloudflare Web Analytics beacon (beacon.min.js) is also embedded. On the first submission the page shows 'Invalid seed. Please re-enter your recovery phrase carefully.' The second submission is sent to the attacker-controlled Vercel domain and the user is redirected to the landing page. ThreatLabz did not observe server-side comparison of the two submissions, so the error appears to be a deception to obtain a second, confirmed phrase.
Impact: possession of the SRP lets the attacker restore the wallet and move all assets without the physical device. Ledger states that its real wallet application never asks users to type the recovery phrase into an online form. No threat actor or malware family is named by the source, no CVE is involved, and no confirmed victim count or loss amount is attributed to this campaign (press reports of large Ledger-related losses concern a separate CryptoBilis matter that researchers have not tied to this site). Zscaler detects the page as HTML.Phish.Ledger. Additional IOCs (several hundred rotating *.vercel.app hosts) are published in the ThreatLabz GitHub IOC repository; the rotating redirect domains and the SRP submission endpoint are not individually listed in the blog IOC table.
MITRE ATT&CK techniques by tactic:
Execution
Defense Evasion
T1497.002 User Activity Based Checks; T1684.001 Impersonation
Exfiltration
T1567 Exfiltration Over Web Service
Resource Development
T1583.006 Web Services; T1583.008 Malvertising
Impact
Remediation
Immediate actions
- Block the published GCS bucket paths, Google Sites pages and Vercel domains, and load the ThreatLabz IOC repository list into proxy/DNS blocklists
- Treat any recovery phrase typed into a web form as compromised: move remaining assets to a new wallet generated from a new recovery phrase using a trusted device
- Alert on web proxy traffic to sites.google.com/view/*ledger* and storage.googleapis.com buckets redirecting to *.vercel.app
Workarounds
- Use an ad blocker or skip sponsored results when searching for wallet software
- Enable TLS-inspecting secure web gateway categories that flag newly observed Vercel-hosted phishing (e.g., Zscaler HTML.Phish.Ledger)
Longer-term hardening
- Train staff and users that Ledger never asks for the 24-word recovery phrase in a browser or app form
- Navigate to hardware wallet vendors via bookmarks or typed URLs instead of sponsored search results
- Deploy browser isolation or ad-blocking for sponsored search results on corporate endpoints
Affected and fixed versions
- Ledger — Ledger hardware wallet users (Ledger Wallet / Ledger Live impersonation)
Vulnerable versions: Users who enter a recovery phrase into the fake site
Timeline
- ThreatLabz observed the fake Ledger page send submitted recovery phrases to an attacker-controlled Vercel domain, showing 'Invalid seed' on the first submission to elicit a second one (August 2026; exact day not stated).
- During the August 2026 observation window, the GCS page redirected to Vercel domains that changed roughly every 15-20 minutes before forwarding to Google Sites pages embedding the phishing iframe (exact day not stated).
- ThreatLabz observed fraudulent sponsored Google ads impersonating Ledger support under a verified advertiser profile (August 2026; exact day not stated in source), targeting users in the US, Europe and parts of Asia.
- ThreatLabz published extended IOCs (several hundred rotating *.vercel.app hosts, GCS buckets, Google Sites pages) in the googleads_ledger_attack GitHub repository referenced by the blog.
- Zscaler ThreatLabz published 'Threat Actors Use Google Ads To Target Ledger Users' with IOCs and the HTML.Phish.Ledger detection name.
- GridinSoft updated its analysis of the campaign, advising that submitted phrases be treated as exposed regardless of the error message.
Sources cited
- Threat Actors Use Google Ads To Target Ledger Users (Zscaler ThreatLabz) zscaler.com · Research · 2026-09-25
- Zscaler Threat Library: HTML.Phish.Ledger threatlibrary.zscaler.com · Vendor report · 2026-09-25
- Malicious Google Ads targeting crypto (SEAL Radar) radar.securityalliance.org · Research
- ThreatLabz IOC repository: googleads_ledger_attack GitHub · Code repository · 2026-09-25
- Ledger Phishing Ads Turn a Fake Device Check Into a Seed-Phrase Trap (GridinSoft) blog.gridinsoft.com · Blog · 2026-10-12
- Fake Ledger Google Ads Scam: How Sponsored Results Steal Recovery Phrases (MalwareTips) malwaretips.com · Blog
- Ledger faces fake website warning amid $86M theft probe (crypto.news) crypto.news · News
Drafted by the Threadlinqs research pipeline from 7 sources; published ; last updated —.