{"id":813,"kind":"story","title":"Safeworld is a startup that evaluates the safety of generative AI robots through simulation","url":"https://techcrunch.com/2026/10/05/can-safeworld-convince-people-that-gen-ai-robots-wont-hurt-them/","host":"techcrunch.com","by":"startupxo","score":0,"comments":0,"time":"2026-10-05T12:44:45.675Z","origin":"curated","language":{"source":"ko","reading":"en","translated":true},"summary":["Safeworld tests robot control software in a simulation environment that includes realistic human behavior models to assess the risks of robots equipped with generative AI.","Unlike traditional algorithms, generative AI is difficult to predict, so Safeworld's co-founders explain that securing safety and reliability before large-scale deployment is crucial.","Safeworld digitally recreates workplace environments, models interactions with robots in various human motion scenarios such as walking or falling, and systematically evaluates risks through thousands of scenarios.","It is expected that third-party validations like Safeworld, alongside evaluations developed internally by robot manufacturers, will play a role in sharing safety standards and enhancing trust across the industry. 3"],"discussion":"https://startupxo.com/en/813/","markdown":"https://startupxo.com/en/813.md","api":"https://startupxo.com/api/v1/item.json?id=813&lang=en","text":null,"edited":null,"sources":[{"site":"techcrunch","name":"TechCrunch","url":"https://techcrunch.com/2026/10/05/can-safeworld-convince-people-that-gen-ai-robots-wont-hurt-them/","score":null,"comments":null,"observed_at":"2026-10-05T12:44:26Z"}],"thread":[]}