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
Physical AI
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.
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.