Can a Robot Do Everything Right and Still Be Unsafe?
What a conversation between Boston Dynamics and VicOne reveals about how manipulated information can turn expected robot operation into an unsafe outcome
Blog
VicOne Situational Awareness Report: 400 Cyber Incidents, with In-Vehicle Systems Most Affected in Q3 2026
VicOne's Q3 2026 Situational Awareness Report tracks where automotive cyber risk shifted: in-vehicle systems overtook enterprise IT as the most affected domain, Asia recorded a 36.8% rise in observed incidents, and ransomware activity spanned the full automotive value chain across 97 incidents.
Sony Patches Seven XAV-9500ES Flaws That Could Have Enabled Device Compromise
Sony patched seven XAV-9500ES vulnerabilities affecting Wi-Fi, Bluetooth, and USB. VicOne examines the attack paths, potential impact, and next steps.
Why Cybersecurity Is Essential to Safety Assurance for Physical AI
Physical AI can become unsafe even without experiencing conventional failures. Safety assurance must account for cyber manipulation of the inputs, models, commands, and systems that shape robot behavior.
AI Agents Can Shape the Physical World. We Can Secure Their Actions.
VicOne supports NVIDIA's Open Agent Safety Platform, extending the focus from secure AI agent decisions to the real-world actions of vehicles and robots.
Looking Beyond the Mythos Era: If AI Can Read Firmware, What Does Protection Mean Now?
Part 2 of 2: AI-driven binary analysis is forcing a rethink of automotive firmware protection. Part 1 covered how AI shortens PoC-to-exploit timelines.
20 Phoenix Contact CHARX Vulnerabilities Patched: How They Could Have Enabled Deeper EV Charger Control
Phoenix Contact patched 20 CHARX SEC-3xxx EV charger vulnerabilities in firmware 1.9.1. VicOne examines their wider implications and practical steps for operators.
CVE-2026-86793: SGLang Bypass Could Let Attackers Run Code on AI Servers
VicOne details CVE-2026-86793, a SafeUnpickler bypass in SGLang that could enable unauthenticated remote code execution, and recommended mitigations.
How Physical AI Challenges Traditional Security Assumptions for Autonomous Systems
Physical AI changes OT security assumptions, making behavioral verification essential when an air gap is not practical and response time is limited.
How BADBOX-Linked Malware Turned Automotive Head Units into Proxy Nodes
VicOne shows how BADBOX-linked malware used DoFun Android head units, weak MQTT security, and trusted OTA updates to push silent, persistent installs.
From Pwn2Own Automotive 2026: Seven Kenwood DNR1007XR Vulnerabilities Now Patched
Kenwood patched seven DNR1007XR vulnerabilities discovered at Pwn2Own Automotive 2026, addressing direct and multistep paths to root access.
FCC, CRA, and IEC 62443: Three Gates to Robotics Market Readiness
Robot manufacturers face distinct FCC Covered List, the EU Cyber Resilience Act (CRA), and IEC 62443 requirements. Learn how shared security workflows can support readiness across all three.
How Mythos Is Reshaping Vulnerability Management: What CISOs Need to Know About VulnOps
Learn how Mythos-ready security and always-on VulnOps help CISOs respond faster to AI-accelerated vulnerabilities, exploit chains, and cyber risk at scale.