Safeworld has emerged from stealth mode, securing $12 million in a seed funding round that concluded around February 9, 2026. The investment round was supported by 515 Ventures, Umami Capital, and Founders Future. Additionally, the company joined the a16z speedrun accelerator, which provided $500 000 in April 2026.
The startup is developing a platform designed to test and evaluate human-robot interactions. By utilizing simulation and generative AI, the company aims to identify rare failure cases in robotic systems before they are deployed in environments where people are present. The firm intends to combine these digital simulations with physical testing protocols.
Based in Oakland and Palo Alto, California, the company currently employs approximately six people. The team includes a professor from Carnegie Mellon University who specializes in robot safety and has received the NSF CAREER Award. Safeworld plans to collaborate with research and open-source communities to establish safety standards for physical AI systems.
