P2P Botnets in the Wild: Pink, Hajime, Mozi, FritzFrog, and Panchan — Decentralized C2 Landscape (360 Netlab Continuous Monitoring)

P2P Botnets in the Wild (TL-2026-0752), also tracked as P2P Botnet Review, is a high-severity malware campaign, first published 2026-06-10. It has no confirmed attribution, affects Huawei HG532 home router / Netgear / ZTE gateways, references 4 CVEs (CVE-2017-17215, CVE-2018-10561, CVE-2018-10562), maps to 25 MITRE ATT&CK techniques (T1021, T1027, T1036), and is covered by 9 detection rules and 23 indicators of compromise.

Key facts for TL-2026-0752

Threat ID
TL-2026-0752
Also known as
P2P Botnet Review, PinkBot, Hajime, Mozi, FritzFrog, Panchan
Severity
HIGH
Status
ACTIVE
Category
MALWARE
First published
2026-06-10
Last reviewed
2026-06-10
Attribution confidence
NONE
Motivation
FINANCIAL
Target sectors
education, government, healthcare, financial, telecommunications, transportation, consumer-iot
Target regions
China, India, Iran, North America, Europe, Global
Detection rules
9
Indicators of compromise
23

Malware and tooling in P2P Botnets in the Wild

Malware and tooling: Gafgyt, XMRig, nbhash

360 Netlab's continuous-monitoring review profiles five active peer-to-peer (P2P) botnet families — Pink, Hajime, Mozi, FritzFrog, and Panchan. Their decentralized command-and-control (no single C2 server) makes takedown difficult. Ranking by size is Pink > Hajime > Mozi >> FritzFrog <> Panchan; FritzFrog and Panchan propagate via SSH weak-credential brute force, while Pink (the largest, peaking at ~1.6M devices) flashes router firmware for persistence.

How P2P Botnets in the Wild works

This intelligence consolidates 360 Netlab's ongoing tracking of the peer-to-peer (P2P) botnet ecosystem alongside the original family-specific research from 360 Netlab, Akamai, NSFOCUS, Kaspersky, and others. P2P botnets replace the classic centralized command-and-control (C2) server with a decentralized overlay network in which every infected node can relay commands, distribute payloads, and propagate the worm. Because there is no single point of failure, sinkholing or seizing one server does not dismantle the botnet, and operators can keep pushing signed configuration updates across the mesh. 360 Netlab tracks these networks with four strategies: active probing (sending crafted probe messages to enumerate peers), traversing recent-communications/peer lists held by DHT nodes, node-heartbeat injection (inserting forged nodes that maintain heartbeats on peer lists), and a 'wait and see' approach that forges DHT identities closer to a target key to capture incoming peer-discovery requests.

Pink is the largest P2P botnet tracked, having infected more than 1.6 million devices (the great majority in China), primarily MIPS-based fiber/broadband routers. NSFOCUS attributed initial compromise to a zero-day against specific broadband-device brands. Pink is a hybrid: it uses a central C2 for time-sensitive commands, a P2P layer for non-urgent instructions, and third-party services (a project hidden on GitHub or Baidu Tieba, plus hard-coded domains delivered over DNS-over-HTTPS, DoH) to bootstrap configuration. It uniquely re-flashes the original firmware of the fiber router after infection — embedding the downloader C2 and a supporting bootloader in the rewritten firmware — to achieve near-permanent control. Pink's purpose is DDoS and HTTP ad injection. A pronounced decline began around 2022-07-12 (dropping to roughly 20,000 daily active nodes), attributed by 360 Netlab to a nationwide cleanup effort by device vendors and C2-blocking actions in late August 2022.

Hajime emerged in October 2016 (months after Mirai). It builds its overlay as a trackerless torrent on top of the public BitTorrent DHT using dynamic info_hashes that rotate daily, and exchanges data with peers over uTorrent's Micro Transport Protocol (uTP). All BitTorrent communications are signed and encrypted using RC4 with private/public keys. Hajime self-propagates via three attack methods: TR-069/CWPM exploitation (the NewNTPServer command-injection vector, typically TCP 7547, sometimes 5555), Telnet default-password attacks, and the Arris cable-modem 'password of the day' attack. The config file carries a signed list of per-architecture 'atk' (attack) and '.i' (infection) modules; nodes search for `.i.xxx` and `atk.xxx` binaries by CPU type. After infection Hajime blocks ports 23, 7547, 5555, and 5358 to lock out Mirai and other competitors, and opens UDP/1457 plus a random high port for DHT/uTP. Infections skew heavily toward MIPS devices, with Iran historically the top country.

Mozi, first identified by 360 Netlab on 2019-12-03, is a DHT-based botnet named after its propagation samples Mozi.m and Mozi.a. It uses a BitTorrent-like DHT to record peer contact information and syncs encrypted configuration files across nodes. Mozi reuses Gafgyt code for common functions (single-instance enforcement, process-name modification, ACL modification) and Gafgyt attack code (HTTP/TCP/UDP DDoS), and later added a crypto-mining component. It spreads through weak/default Telnet credentials and unpatched CVEs including CVE-2017-17215 (Huawei HG532), CVE-2018-10561/CVE-2018-10562 (GPON home routers), and CVE-2014-8361 (Realtek SDK miniigd UPnP). Targets expanded to Netgear, Huawei, and ZTE network gateways; infections concentrated in India and China.

FritzFrog, first observed 2020-01-09, is a Golang, modular, multi-threaded, fileless worm with a completely proprietary P2P protocol written from scratch (no μTP/known protocol). It brute-forces SSH credentials against government, education, medical, banking, and telecom targets (500+ servers breached, including U.S./European universities and a railway company). On success it installs an RSA public key into the victim's authorized_keys for passwordless re-entry, runs in memory as processes named 'ifconfig' and 'nginx', and listens on TCP 1234 for commands. Payloads — including a 'libexec' XMRig-based Monero miner pointed at web.xmrpool.eu:5555 — are distributed as in-memory blobs split with hash tracking and retrieved via P2P 'getbin' or HTTPS to port 1234. The protocol uses AES symmetric encryption with Diffie–Hellman key exchange over JSON, implements 30+ commands, is UPX-packed, and has shipped in 20+ versions; nodes vote to distribute brute-force targets so no two attack the same host. Akamai noted a wallet relationship linking FritzFrog to Mozi.

Panchan, discovered 2022-03-19, is a Go (compiled 1.17.7–1.18) SSH worm and cryptominer. It propagates by SSH dictionary attacks (default creds such as root, ubuntu, debian, pi, user) and by harvesting SSH keys from ~/.ssh/id_rsa and known_hosts for lateral movement. Persistence is via a systemd service 'systemd-worker' with the binary copied to /bin/systemd-worker and a renamed 'xinetd' copy in a randomly named hidden folder under /. It opens TCP 1919 (plaintext P2P, firewalled open via iptables); peer messages start with 'pan-chan's mining rig hi!' and end with 'finish'. It runs xmrig and nbhash miners fileless via memfd_create (base64-decoded, memory-mapped), with NiceHash pool traffic over ports 3380/3387, and reports infections via a Discord webhook. Akamai found 209 peers (40 active), with education the top vertical, and the admin panel contained Japanese (katakana) text suggesting Japanese-fluent developers.

Defensive priorities: enforce strong, unique SSH credentials and key-only auth; monitor for the family-specific process names, ports, and persistence artifacts above; patch the Mozi CVEs on exposed IoT/routers; segment IoT/OT networks; and treat decentralized P2P C2 as resistant to single-point takedown — blocking and host-level eradication, not sinkholing, is the effective control.

MITRE ATT&CK techniques used in TL-2026-0752

Lateral Movement

T1021 Remote Services

Defense Evasion

T1027 Obfuscated Files or Information; T1036 Masquerading; T1564 Hide Artifacts; T1620 Reflective Code Loading

Discovery

T1046 Network Service Discovery

Execution

T1059 Command and Scripting Interpreter

Command and Control

T1071 Application Layer Protocol; T1090 Proxy; T1102 Web Service; T1105 Ingress Tool Transfer; T1571 Non-Standard Port; T1573 Encrypted Channel

Initial Access

T1078 Valid Accounts; T1133 External Remote Services; T1190 Exploit Public-Facing Application

Persistence

T1098 Account Manipulation; T1542 Pre-OS Boot; T1543 Create or Modify System Process

Credential Access

T1110 Brute Force; T1552 Unsecured Credentials

Impact

T1496 Resource Hijacking; T1498 Network Denial of Service

Resource Development

T1584 Compromise Infrastructure

defense-impairment

T1685 Disable or Modify Tools

Affected products and versions in P2P Botnets in the Wild

  • Huawei — HG532 home router / Netgear / ZTE gateways
    Vulnerable versions: HG532 (CVE-2017-17215); unpatched gateways
    Fixed in: vendor-patched firmware
  • Realtek — Realtek SDK (miniigd UPnP)
    Vulnerable versions: devices using vulnerable Realtek SDK (CVE-2014-8361)
    Fixed in: patched SDK firmware
  • Various (GPON) — GPON home routers
    Vulnerable versions: CVE-2018-10561 / CVE-2018-10562 affected GPON ONTs
    Fixed in: patched firmware
  • Generic — Linux servers exposing SSH; MIPS-based IoT/fiber routers
    Vulnerable versions: weak/default SSH credentials; weak/default Telnet credentials
    Fixed in: key-only SSH, strong credentials

Remediation for P2P Botnets in the Wild

Patches

  • Patch CVE-2017-17215 (Huawei HG532), CVE-2018-10561 and CVE-2018-10562 (GPON routers), and CVE-2014-8361 (Realtek SDK miniigd UPnP) on exposed IoT/router fleets to deny Mozi propagation.
  • Apply vendor firmware updates for Netgear, Huawei, and ZTE gateways targeted by Mozi.

Immediate actions

  • Enforce strong, unique SSH passwords and prefer key-only authentication; disable password auth on internet-facing SSH (mitigates FritzFrog and Panchan brute force).
  • Audit ~/.ssh/authorized_keys for unrecognized RSA keys (FritzFrog backdoor) and known_hosts/id_rsa exposure (Panchan key harvesting).
  • Hunt for malicious process names: ifconfig/nginx/libexec (FritzFrog), systemd-worker/xinetd (Panchan), and listeners on TCP 1234 (FritzFrog) and 1919 (Panchan).
  • Block outbound traffic to web.xmrpool.eu and known mining pool ports (5555, 3380, 3387).

Workarounds

  • Disable or firewall Telnet (TCP 23) and remote management on IoT devices.
  • Restrict SSH exposure to VPN/bastion and apply fail2ban-style rate limiting and account lockout.

Longer-term hardening

  • Deploy EDR/host monitoring with behavioral detection for fileless execution (memfd_create) and P2P beaconing.
  • Segment and isolate IoT/OT and broadband-router management planes; restrict TR-069/CWMP (TCP 7547) to trusted ISP subnets.
  • Maintain firmware integrity monitoring on fiber/broadband routers (Pink re-flashes firmware for persistence).
  • Continuously monitor DHT/BitTorrent-like traffic patterns and non-standard P2P ports egressing the environment.

CVEs associated with P2P Botnets in the Wild

CVE-2017-17215, CVE-2018-10561, CVE-2018-10562, CVE-2014-8361

Weaknesses (CWE) in P2P Botnets in the Wild

CWE-1392, CWE-307, CWE-78, CWE-798

Timeline of P2P Botnets in the Wild

  • Hajime IoT botnet emerges, months after Mirai, building a trackerless-torrent C2 over the public BitTorrent DHT with daily-rotating info_hashes (Kaspersky/Securelist).
  • 360 Netlab first identifies Mozi, a DHT-based P2P botnet named after its propagation samples Mozi.m / Mozi.a, reusing Gafgyt code.
  • FritzFrog first observed — Golang, fileless, proprietary-P2P SSH brute-force worm; Akamai/Guardicore later documents 500+ breached servers.
  • Mozi reported expanding to Netgear, Huawei, and ZTE network gateways, becoming a dominant source of malicious IoT traffic.
  • 360 Netlab and NSFOCUS disclose Pink — the largest tracked P2P botnet at 1.6M+ infected MIPS fiber routers, using GitHub/Baidu Tieba + DoH config and firmware flashing for persistence.
  • Panchan, a Go SSH-worm cryptominer, discovered by Akamai; 209 peers (40 active), education the top vertical, Japanese strings in admin panel.
  • Pink begins a pronounced decline, dropping toward ~20,000 daily active nodes, attributed to a nationwide vendor cleanup and C2-blocking actions through late August 2022.
  • 360 Netlab publishes the continuous-monitoring review ranking the active P2P families: Pink > Hajime > Mozi >> FritzFrog <> Panchan.

Sources cited for P2P Botnets in the Wild

Threats related to P2P Botnets in the Wild

Detection coverage for TL-2026-0752

As of 2026-06-10, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-0752 across Splunk SPL, Microsoft KQL and Sigma, covering 23 indicator(s) of compromise. The whole corpus is readable without an account; a free account unlocks full detection query text in Splunk SPL, Microsoft KQL and Sigma; paid tiers add raw indicator values, correlation and the MCP server. Threadlinqs MCP server · View plans.

Threadlinqs Intelligence — Real-Time Threat Detection Platform

[ 0 threats ] [ 0 det ] [ CRIT: 0 ] [ HIGH: 0 ]
// threat_feed
$ sort --newest
Showing all threats

Latest Threats