home bbs files messages ]

Just a sample of the Echomail archive

Cooperative anarchy at its finest, still active today. Darkrealms is the Zone 1 Hub.

   CONSPRCY      How big is your tinfoil hat?      2,445 messages   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]

   Message 1,819 of 2,445   
   Mike Powell to All   
   UniPwn exposes Unitree vu   
   08 Oct 25 08:56:22   
   
   TZUTC: -0500   
   MSGID: 1568.consprcy@1:2320/105 2d4ba9a6   
   PID: Synchronet 3.21a-Linux master/123f2d28a Jul 12 2025 GCC 12.2.0   
   TID: SBBSecho 3.28-Linux master/123f2d28a Jul 12 2025 GCC 12.2.0   
   BBSID: CAPCITY2   
   CHRS: ASCII 1   
   FORMAT: flowed   
   Calm down everyone - Unitree's recently discovered exploit will absolutely,   
   definitely not give rise to the first robot-to-robot viral infection using   
   Bluetooth   
      
   Date:   
   Tue, 07 Oct 2025 20:32:00 +0000   
      
   Description:   
   UniPwn exposes chain of Unitree robot vulnerabilities, allowing root-level   
   command execution and potential wireless propagation between devices.   
      
   FULL STORY   
      
   Security researchers Bin4ry and d0tslash have published a write-up on GitHub   
   about an exploit named "UniPwn" which affects multiple Unitree product lines.    
      
   The vulnerability affects G1 humanoids, Go2, and B2 quadrupeds, and it can be   
   used to escalate privileges to root.    
      
   It appears to chain together weaknesses that, when combined, permit remote   
   command injection on affected devices.   
      
   How the vulnerability works and why it matters   
      
   The vulnerability set reportedly includes hardcoded cryptographic keys and a   
   handshake that checks only for the string "unitree", and also includes   
   unsanitized user data concatenated into shell commands the system runs.    
      
   Those elements combine into an unusually straightforward path from a network   
   packet to arbitrary code execution.    
      
   Because the exposed service accepts wireless connections, a compromised unit   
   can receive commands and attempt to influence devices within radio range.    
      
   That changes the threat model from a single exploited device to potential   
   lateral movement across nearby units.    
      
   The researchers say the exploit leverages a Bluetooth Low Energy and Wi-Fi   
   configuration service.    
      
   This means a compromised unit can receive commands over wireless links and   
   potentially attempt to influence devices within radio range.    
      
   The researchers describe parts of the UniPwn chain as "wormable", meaning   
   successful exploitation can allow malicious code to persist and attempt   
   propagation, which raises the risk because it could permit automated spread   
   between reachable devices.    
      
   Yet wormable behavior observed in tests does not guarantee rapid real-world   
   propagation.    
      
   Real-world spread depends on device configuration, network segmentation,   
   firmware diversity, physical proximity, vendor patching pace, and operator   
   practices.    
      
   Controlled lab tests can show a capability, but field propagation will be   
   shaped by those operational factors.    
      
   Thus, this first robot-to-robot viral infection remains unlikely, although   
   manufacturers and operators would be unwise to treat this as a remote   
   theoretical threat.    
      
   Independent research into jailbreaking LLM-powered robots increases the   
   urgency of these technical findings.    
      
   A project known as RoboPAIR demonstrated that carefully crafted prompts can   
   coerce robot controllers, including the Unitree Go2, to perform harmful   
   actions.    
      
   Reported scenarios include converting robots into covert surveillance   
   platforms and guiding them to place explosives.    
      
   The RoboPAIR team reported high success rates when it supplied the target   
   robots API and formatted prompts that the API executed as code.    
      
   Combining LLM jailbreak techniques with low-level remote command injection   
   expands the attack surface.    
      
   This is because a single compromise could both defeat model safeguards and   
   execute arbitrary system commands.    
      
   Therefore, this disclosure should prompt immediate mitigation efforts,    
   clearer vendor communication, and realistic threat modeling to avoid   
   preventable harm.    
      
   The nature of this flaw is technically notable, and if weaponized, the   
   consequences could be severe.    
      
   Via Toms Hardware    
      
   ======================================================================   
   Link to news story:   
   https://www.techradar.com/pro/calm-down-everyone-unitrees-recently-discovered-   
   exploit-will-absolutely-definitely-not-give-rise-to-the-first-robot-to-robot-v   
   iral-infection-using-bluetooth   
      
   $$   
   --- SBBSecho 3.28-Linux   
    * Origin: capitolcityonline.net * Telnet/SSH:2022/HTTP (1:2320/105)   
   SEEN-BY: 105/81 106/201 128/187 129/14 305 153/7715 154/110 218/700   
   SEEN-BY: 226/30 227/114 229/110 111 206 300 307 317 400 426 428 470   
   SEEN-BY: 229/664 700 705 266/512 291/111 320/219 322/757 342/200 396/45   
   SEEN-BY: 460/58 633/280 712/848 902/26 2320/0 105 304 3634/12 5075/35   
   PATH: 2320/105 229/426   
      

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]


(c) 1994,  bbs@darkrealms.ca