Cybersecurity AI: Hacking Consumer Robots in the AI Era
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908877504643072 |
|---|---|
| author | Mayoral-Vilches, Víctor Ayucar-Carbajo, Unai Laflamme, Olivier Peng, Ruikai Sanz-Gómez, María Balassone, Francesco Apa, Lucas Gil-Uriarte, Endika |
| author_facet | Mayoral-Vilches, Víctor Ayucar-Carbajo, Unai Laflamme, Olivier Peng, Ruikai Sanz-Gómez, María Balassone, Francesco Apa, Lucas Gil-Uriarte, Endika |
| contents | Is robot cybersecurity broken by AI? Consumer robots -- from autonomous lawnmowers to powered exoskeletons and window cleaners -- are rapidly entering homes and workplaces, yet their security remains rooted in assumptions of specialized attacker expertise. This paper presents evidence that Generative AI has fundamentally disrupted robot cybersecurity: what historically required deep knowledge of ROS, ROS 2, and robotic system internals can now be automated by anyone with access to state-of-the-art GenAI tools spearheaded by the open source CAI (Cybersecurity AI). We provide empirical evidence through three case studies: (1) compromising a Hookii autonomous lawnmower robot, uncovering fleet-wide vulnerabilities and data protection violations affecting 267+ connected devices, (2) exploiting a Hypershell powered exoskeleton, demonstrating safety-critical motor control weaknesses and credential exposure including access to over 3,300 internal support emails, and (3) breaching a HOBOT S7 Pro window cleaning robot, achieving unauthenticated BLE command injection and OTA firmware exploitation. Across these platforms, CAI discovered in an automated manner 38 vulnerabilities that would have previously required months of specialized security research. Our findings reveal a stark asymmetry: while offensive capabilities have been democratized through AI, defensive measures often remain lagging behind. We argue that traditional defense-in-depth architectures like the Robot Immune System (RIS) must evolve toward GenAI-native defensive agents capable of matching the speed and adaptability of AI-powered attacks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_08665 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Cybersecurity AI: Hacking Consumer Robots in the AI Era Mayoral-Vilches, Víctor Ayucar-Carbajo, Unai Laflamme, Olivier Peng, Ruikai Sanz-Gómez, María Balassone, Francesco Apa, Lucas Gil-Uriarte, Endika Cryptography and Security Is robot cybersecurity broken by AI? Consumer robots -- from autonomous lawnmowers to powered exoskeletons and window cleaners -- are rapidly entering homes and workplaces, yet their security remains rooted in assumptions of specialized attacker expertise. This paper presents evidence that Generative AI has fundamentally disrupted robot cybersecurity: what historically required deep knowledge of ROS, ROS 2, and robotic system internals can now be automated by anyone with access to state-of-the-art GenAI tools spearheaded by the open source CAI (Cybersecurity AI). We provide empirical evidence through three case studies: (1) compromising a Hookii autonomous lawnmower robot, uncovering fleet-wide vulnerabilities and data protection violations affecting 267+ connected devices, (2) exploiting a Hypershell powered exoskeleton, demonstrating safety-critical motor control weaknesses and credential exposure including access to over 3,300 internal support emails, and (3) breaching a HOBOT S7 Pro window cleaning robot, achieving unauthenticated BLE command injection and OTA firmware exploitation. Across these platforms, CAI discovered in an automated manner 38 vulnerabilities that would have previously required months of specialized security research. Our findings reveal a stark asymmetry: while offensive capabilities have been democratized through AI, defensive measures often remain lagging behind. We argue that traditional defense-in-depth architectures like the Robot Immune System (RIS) must evolve toward GenAI-native defensive agents capable of matching the speed and adaptability of AI-powered attacks. |
| title | Cybersecurity AI: Hacking Consumer Robots in the AI Era |
| topic | Cryptography and Security |
| url | https://arxiv.org/abs/2603.08665 |