This Is Why Humanoid Robots Still Cannot Be Fully Trusted

Human-like

robots are the big revelation of the tech industry after the chatbots and AI agents. Imagine walking into a cafe or a factory and finding out that a robot is serving people and manufacturing products in a factory. But as these machines become a key part of our lives, an important question remains: can they work safely around people?

Humanoid Robots Are Advancing Sooner Than Expected

According to the Wall Street Journal, humanoid robots impressed visitors by delivering snacks, shaking hands, and even dancing at the Automate conference in Chicago. This event highlighted how quickly the humanoid robots are evolving, with companies hoping these robots will eventually work in factories, warehouses and even homes. However, recent viral videos have also raised concerns. Reportedly, a humanoid robot appeared to dance uncontrollably inside a restaurant, while another was seen kicking a child during a performance in China. Robot makers have often said there have been no known cases of a humanoid robot seriously injuring or killing a person, the incidents have renewed discussions around safety. Michele Silva, a functional safety expert at Reynolds & Moore, told WSJ that today’s humanoid robots weigh close to 200 pounds (nearly 90kg). He flags that if a robot suddenly loses power or balance and falls on a human, it can crush them and hurt badly. The flexibility that makes humanoids useful also makes them harder to predict.

AI Needs Its Own Safety System

Companies developing humanoid robots maintain their stance that safety has become one of their biggest priorities. Chipmaker Nvidia recently introduced a new safety system built around its Blackwell chips. As per Amit Goel, Senior Director of Robotics and Edge AI Ecosystem at Nvidia, the technology constantly checks information from cameras and sensors to identify hazards. As humanoid robots become more capable, global safety rules are also being developed. Reportedly, an expert panel at the International Organization for Standardization (ISO) is currently working on safety guidelines for humanoid robots, with a new international standard expected by 2028. The race to build humanoid robots is no longer just about creating smarter machines. Rather it is about earning people's trust. Whether these robots end up working beside us in factories, hospitals or even our homes will depend not only on what they can do, but on how safely they can do it.