Artificial Intelligence

Researchers Identify 'UniPwn' Cybersecurity Flaw in Unitree Robots

Published SEP 30 2025

Story

Above: Unitree humanoid robot at the World Smart Industry Expo in Chongqing, China on Sept. 7, 2025. Image credit: He Penglei/China News Service/VCG/Getty Images

The Facts

Security researchers have reportedly discovered a critical vulnerability called "UniPwn" that affects Unitree's Go2 and B2 quadruped robots and G1 and H1 humanoid models. The flaw allegedly exploits hardcoded encryption keys in the Bluetooth Low Energy provisioning protocol.
arXiv
IEEE Spectrum
The vulnerability reportedly allows attackers to gain root access by injecting malicious code disguised as Wi-Fi credentials. Once compromised, robots can automatically scan for and infect nearby Unitree robots via Bluetooth.
Pepelac News
ZME Science
The researchers also claim that the Unitree G1 humanoid robot continuously transmits sensor data, audio, video and GPS coordinates to Chinese servers every 300 seconds — without user notification or consent mechanisms.
arXiv
Decrypt
Unitree’s proprietary FMX encryption system is said to use static Blowfish-ECB encryption and predictable random number generation, enabling offline decryption of configuration files and firmware without brute force attacks.
arXiv
Decrypt
Security researchers Andreas Makris and Kevin Finisterre first contacted Unitree in May to report the vulnerabilities. However, according to the researchers, the company stopped responding in July, prompting the public disclosure of the flaws.
Pepelac News
IEEE Spectrum
The study reportedly demonstrated that a cybersecurity AI agent running on the G1 robot could autonomously conduct reconnaissance and prepare offensive cyber operations, effectively transforming the humanoid into an active attack platform.
Decrypt
ZME Science

Sources Split

LEFT

CENTER

RIGHT

ESTAB
CRIT

CENTER

PRO
ESTAB

The Spin

Techno-skeptic narrative

Unitree’s humanoids aren’t just robots — they’re Wi Fi–connected spies in secret contact with China. They ship with factory-set keys and a Bluetooth flaw that lets attackers inject commands, gain root, worm to nearby units, and run a built-in AI to map networks. Despite receiving warnings from security researchers in May, Unitree has left these toy-like humanoids — including those used by China’s PLA — to function as covert surveillance devices and fully weaponizable cyber platforms.

arXiv

Mario Nawfal

Techno-optimist narrative

With no American humanoids available, Chinese robots are the only option for developers to build advanced robotics software — and Unitree is proving they can be trusted. The company has quickly addressed recent vulnerabilities, rolled out fixes, and strengthened permission management, keeping robots offline by default and connecting only with user authorization. Far from being spy platforms, these robots are now secure, privacy-conscious, and fully under user control.

Unitree on X

Cix Liv on X

Metaculus Prediction

Editor's Note

This story currently has limited coverage. We will continue to monitor all major outlets and update our reporting as additional information becomes available.