# BADBOX 3.0 Android TV Box Malware Drives ~40 Billion Fraudulent Ad Bid Requests Daily

> BADBOX 3.0 is an ad-fraud and residential-proxy operation running on low-cost AOSP Android streaming boxes that are not Play Protect certified. HUMAN Security's Satori team reports roughly 40 billion bid requests per day at peak (over 4% of observed CTV traffic) and links the software catalog to the Chinese manufacturer TXCZ, with Vo1d botnet code found alongside the new fraud modules.

- **Published:** 2026-11-19T14:14:55Z
- **Last reviewed:** 2026-11-19T14:14:55Z
- **Canonical:** https://intel.threadlinqs.com/threat/TL-2026-6462
- **Card image:** https://intel.threadlinqs.com/og/threat/TL-2026-6462.png
- **ID:** TL-2026-6462
- **Severity:** HIGH
- **Category:** MALWARE
- **Status:** ACTIVE
- **Actor:** BADBOX operators (China)
- **Detections:** 9 (1 shown free below) · **IOCs:** 29 (3 shown free below)

## Description

HUMAN Security's Satori Threat Intelligence team published research on 2026-11-18 describing BADBOX 3.0, the successor to BADBOX and BADBOX 2.0. It was reported by Cyber Security News on 2026-11-19. The operation abuses low-cost Android streaming and TV boxes that run the Android Open Source Project (AOSP) rather than Google's Android TV OS and are not Play Protect certified. At peak the fraud produced about 40 billion bid requests per day, over 4% of the connected TV (CTV) traffic HUMAN observes and more than three times the daily volume of VASTFLUX, the previous record holder. These figures count bid requests, not ads watched, payments or infected devices. The source gives no unique infected-device count for BADBOX 3.0.

Infection starts with a backdoor packaged as a system app. It contacts a command-and-control server, which tells the box which modules to download. The modules come from a maintained collection of ad-fraud and proxy components. Backup copies sit on separate hosting services, so taking down one copy may not stop new infections. In one analyzed sample, a system app plays video ads in a transparent player placed off-screen, so the owner sees nothing change on the TV. Bid requests spoof the names of legitimate CTV apps that do not run on the infected boxes. Some of the fraud needs no real device at all: some requests falsely claim a server inserted an ad into a viewer's stream when no device or ad playback existed.

The malware can also enroll a box in any of 11 residential proxy networks without the owner's consent. Satori assessed PXYEDGE to be a renamed successor to IpMoYu. For one impersonated app, about 80% of U.S. requests came from addresses flagged as residential proxy exits. Satori tied the BADBOX 3.0 software catalog to the Chinese manufacturer TXCZ through server infrastructure, a factory-installed launcher, and code that checks for TXCZ-specific device settings. Code from the Vo1d Android TV botnet appeared next to the new fraud modules, strengthening a link noted in earlier BADBOX research.

Satori's detection signals: repeated traffic spikes at nearly the same hour, a browser identifier from an internal test build, and impersonated-app requests with a high share of residential proxy exit addresses. Defenders should not block an impersonated app name wholesale, because Satori saw no malicious behavior when it tested one such app on its intended hardware. A box that streams normally may still be compromised. Note: the primary HUMAN blog post could not be fetched during this research (HTTP 403), so BADBOX 3.0 specifics here rest on the Cyber Security News report. No CVE or CVSS score is stated in the source, and package names and file hashes were not published.

## MITRE ATT&CK

- T1195.003 Compromise Hardware Supply Chain
- T1583.001 Domains
- T1071.001 Web Protocols
- T1090 Proxy
- T1573 Encrypted Channel
- T1036 Masquerading
- T1496 Resource Hijacking

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

**BADBOX 3.0 Backdoor and Module C2 DNS Lookups (drgdj.top, pixelpioneerss.com, thedynamicleap.com) — Exploitation** (`TL-2026-6462-SPL-001`)

TL-2026-6462: DNS resolution of the BADBOX 3.0 backdoor parent domain drgdj.top (api/file/gh subdomains) and the module C2 servers api.pixelpioneerss.com and bur.thedynamicleap.com. A pre-installed system-app backdoor on a non-Play-Protect-certified AOSP streaming box contacts these hosts to fetch its module list, so any hit from a consumer or IoT segment is a strong indicator of a compromised TV box.

- **Languages:** Splunk SPL, Microsoft KQL, Sigma
- **ATT&CK:** T1195.003, T1071.001, T1583.001
- **Data sources:** dns_logs, stream:dns, DnsEvents
- **Severity:** high
- **Confidence:** high

### Splunk SPL

```spl
index=dns sourcetype="stream:dns" earliest=-24h
  (query="drgdj.top" OR query="*.drgdj.top" OR query="pixelpioneerss.com" OR query="*.pixelpioneerss.com" OR query="thedynamicleap.com" OR query="*.thedynamicleap.com")
| stats count AS lookups dc(query) AS distinct_names values(query) AS queried_names min(_time) AS first_seen max(_time) AS last_seen BY src_ip
| eval first_seen=strftime(first_seen,"%F %T"), last_seen=strftime(last_seen,"%F %T")
| table src_ip lookups distinct_names queried_names first_seen last_seen
```

### Microsoft KQL

```kql
DnsEvents
| where TimeGenerated > ago(24h)
| extend QueryName = tolower(trim_end(@"\.", Name))
| where QueryName matches regex @"(^|\.)(drgdj\.top|pixelpioneerss\.com|thedynamicleap\.com)$"
| summarize Lookups = count(), QueryNames = make_set(QueryName, 20), FirstSeen = min(TimeGenerated), LastSeen = max(TimeGenerated) by ClientIP
| project ClientIP, Lookups, QueryNames, FirstSeen, LastSeen
```

### Sigma

```yaml
title: BADBOX 3.0 Backdoor and Module C2 DNS Lookups
id: 3b6d1f0e-6462-4a11-9c01-5a7e2d9c0001
status: experimental
description: Detects DNS queries for BADBOX 3.0 backdoor and module C2 domains (drgdj.top, pixelpioneerss.com, thedynamicleap.com) used by TXCZ-linked AOSP TV box malware. TL-2026-6462.
references:
    - https://cybersecuritynews.com/badbox-3-0-malware/
    - https://www.fbi.gov/investigate/cyber/alerts/2025/home-internet-connected-devices-facilitate-criminal-activity
author: Threadlinqs Team
date: 2026/11/19
tags:
    - attack.initial-access
    - attack.resource-development
    - attack.command-and-control
    - attack.t1195.003
    - attack.t1071.001
    - attack.t1583.001
logsource:
    category: dns
detection:
    selection_sub:
        query|endswith:
            - '.drgdj.top'
            - '.pixelpioneerss.com'
            - '.thedynamicleap.com'
    selection_exact:
        query:
            - 'drgdj.top'
            - 'pixelpioneerss.com'
            - 'thedynamicleap.com'
    condition: 1 of selection_*
falsepositives:
    - Threat-intel research or sandbox hosts deliberately resolving BADBOX infrastructure
    - Security-vendor feeds or DNS sinkhole tests that query the listed domains
level: high
```

**Known false positives:**

- Threat-intel research or sandbox hosts deliberately resolving BADBOX infrastructure
- Security-vendor feeds or DNS sinkhole tests that query the listed domains

**Tuning notes:** Point the index/sourcetype at your DNS source (Zeek bro_dns, Infoblox, Windows DNS, Splunk Stream). Strip trailing dots from query if your source keeps them. Scope to guest/IoT VLANs to cut noise and add known research hosts to an exclusion lookup. Treat any hit as a compromised device: isolate the box rather than only blocking the name.

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

- BADBOX 3.0 Legacy C2 Endpoint Connection [indicator]:8081 — Exploitation (locked: unlock with TL Creds)
- BADBOX 3.0 Ad-Fraud Tasking Servers ([indicator], [indicator], [indicator], speed25.top) — Post-Exploitation (locked: unlock with TL Creds)
- BADBOX 3.0 Hour-Aligned Beaconing from IoT/CTV Segment — C2 / Behavioral (locked: unlock with TL Creds)
- BADBOX 3.0 Module-List and Payload Host DNS (api/file/[indicator]) — Exploitation (locked: unlock with TL Creds)
- BADBOX 3.0 Encrypted Secondary Payload Downloader IPs ([indicator]:8801, [indicator]:8802) — Post-Exploitation (locked: unlock with TL Creds)
- BADBOX 3.0 PXYEDGE / IpMoYu Residential Proxy Domain DNS ([indicator], [indicator]) — C2 / Proxy (locked: unlock with TL Creds)
- BADBOX 3.0 Hidden-Ad Config and Payload Staging DNS (shzhtxcz.com, Aliyun OSS terrtest) — Post-Exploitation (locked: unlock with TL Creds)
- BADBOX 3.0 PXYEDGE Residential Proxy Server Connections — C2 / Proxy (locked: unlock with TL Creds)

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

## Sample indicators (3 of 29)

Defanged, TLP:CLEAR.

- IP address: `47.113.90[.]237`
- Domain: `drgdj[.]top`
- IP address: `39.108.226[.]196`

Full set of 29 indicators unlocks with TL Creds: https://threadlinqs.com/tl-creds

## Sources

- [BADBOX 3.0 Malware Infects Android TV Boxes to Generate 40 Billion Fraudulent Ad Requests Daily](https://cybersecuritynews.com/badbox-3-0-malware/)
- [HUMAN Satori: BADBOX 3.0 hides the largest ad fraud operation ever disclosed inside infected TV boxes](https://www.humansecurity.com/learn/blog/satori-threat-intelligence-disruption-badbox-3-0-hides-the-largest-ad-fraud-operation-ever-disclosed-inside-infected-tv-boxes/)
- [FBI: Home Internet-Connected Devices Facilitate Criminal Activity (BADBOX advisory)](https://www.fbi.gov/investigate/cyber/alerts/2025/home-internet-connected-devices-facilitate-criminal-activity)
- [Google taking legal action against the BADBOX 2.0 botnet](https://blog.google/innovation-and-ai/technology/safety-security/google-taking-legal-action-against-the-badbox-20-botnet/)
- [HUMAN Satori: Disrupting BADBOX 2.0](https://www.humansecurity.com/learn/blog/satori-disrupting-badbox-2/)
- [BADBOX 2.0 Botnet Infects 1 Million Devices (The Hacker News)](https://thehackernews.com/2025/03/badbox-20-botnet-infects-1-million.html)
- [CISO who helped unmask Badbox warns: Version 3 is coming (The Register)](https://www.theregister.com/2025/06/11/badbox_round_three/)
- [Google sued BADBOX 2.0 malware botnet operators (Cyber Security News)](https://cybersecuritynews.com/google-sued-badbox-2-0-malware-botnet-operators/)
- [1.6 million Android TVs worldwide hacked - Vo1d (Cyber Security News)](https://cybersecuritynews.com/1-6-million-android-tvs-worldwide-hacked/)
- [Trojans All the Way Down: BADBOX and PEACHPIT (HUMAN)](https://www.humansecurity.com/learn/blog/trojans-all-the-way-down-badbox-and-peachpit/)

## 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-2026-6462
