VicOne Showcases Physical AI Cybersecurity at NVIDIA GTC Berlin

Demonstrations show how robotics teams can test behavior-changing cyber risks before deployment and respond during operation.

MUNICH, MUNICH, GERMANY, October 23, 2026 /EINPresswire.com/ — VicOne, a Japan-headquartered leader in physical AI cybersecurity, today shared two demonstrations at NVIDIA GTC Berlin. The demonstrations showed how robotics teams can test cyber risks that change robot behavior before deployment and assess and respond to related threats during operation.

A robot can be misled into unsafe behavior when the information guiding its actions has been manipulated. Cyber threats add an adversary to the safety equation, making hazardous scenarios something that can be deliberately triggered. Cybersecurity strengthens safety assurance by helping teams examine whether safeguards remain effective under manipulation and detect related changes during operation.

The showcase brought together NVIDIA technologies, and VicOne’s Physical AI cybersecurity capabilities. One demonstration used NVIDIA Nemotron™ models to support contextual reasoning within VicOne Rthena, which assesses runtime risk and coordinates approved responses across a robot fleet. The other used the VicOne Radeis Extension with NVIDIA Isaac Sim™ to test how cyber manipulation could change robot behavior before deployment.

Together, the demonstrations highlighted how cybersecurity helps keep robots trustworthy as AI decisions become physical actions.

“Robot makers need to know whether the safeguards they design remain effective when the information guiding a robot is deliberately manipulated,” said Max Cheng, Chief Executive Officer of VicOne. “By testing those conditions before deployment and detecting related threats during operation, we help teams strengthen the evidence supporting robot safety.”

Detecting and Mitigating Runtime Threats Before They Impact Robot Behavior

The runtime demonstration featured a simulated hospital service scenario in which a robot transporting medication near a pediatric ward encountered suspicious commands and anomalous communication signals that could potentially influence its navigation and operational behavior.

VicOne uses AI to protect AI, with NVIDIA Nemotron serving as the foundation for training Rthena’s AI cybersecurity model and bringing protection to the robot’s edge. Rthena continuously monitored runtime activities to detect malicious intent and assess potential behavioral risks.

Upon detecting a threat, Rthena could recommend or apply adaptive protection strategies, such as real-time prompt patching and policy enforcement, to neutralize malicious instructions before they affected the robot’s decisions or actions. This demonstrated how Rthena could prevent malicious instructions from affecting robot behavior while maintaining operational safety and minimizing disruption to the robot’s mission.

Testing the Effects of Cyber-Safety Manipulation Before Deployment

The second demonstration showed how an adversarial visual input could redirect a robot’s attention, causing it to ignore a stop sign and continue moving.

Using the Radeis Extension with NVIDIA Isaac Sim, teams could examine the resulting behavior change in simulation. This supported investigation of how manipulated inputs affect safeguards and the collection of evidence before physical testing and release.

For robotics teams serving Europe, the EU Machinery Regulation, which applies from January 20, 2027, includes requirements to protect safety-relevant systems and data against accidental or intentional corruption. VicOne helps teams assess related cyber risks and build evidence relevant to their robot’s design and operating conditions.

VicOne’s free Radeis Extension for NVIDIA Isaac Sim is available now. The broader Radeis and Rthena platforms are commercially available.

About VicOne
VicOne, headquartered in Tokyo, Japan, is a cybersecurity leader for Physical AI, built on proven automotive cybersecurity expertise. Its software and services protect the digital systems that shape how vehicles and robots see, decide, and act. VicOne helps OEMs, Tier 1 suppliers, robot makers, and operators identify cyber risks early, assess their potential safety impact, and protect systems in operation, helping maintain safe, reliable behavior across real-world deployments. With 180+ zero-days uncovered across automotive and robotics, 100M+ new threat intelligence signals added monthly, and 30 U.S. patent applications pending in AIDV and AI security, VicOne combines specialized Physical AI security research with global threat intelligence. As a Trend Micro subsidiary, VicOne is backed by decades of global cybersecurity expertise. For more information, visit vicone.com.

Ling Cheng
VicOne
344002265 ext.
Visit us on social media:
LinkedIn
YouTube

VicOne Showcases Physical AI Cybersecurity at NVIDIA GTC Berlin

Media gallery