The biggest trend in robotics is handing control over to generative AI models—but that introduces a critical problem: generative architectures aren't predictable the way traditional algorithms are. How can you be sure your brand-new humanoid robot will be safe?
Dr. Ding Zhao, who directs the Safe AI Lab at Carnegie Mellon University, has spent nearly his entire career on this problem. Now, alongside veteran startup executive Kyle Wong and machine learning engineer Simo Rachidi, he has founded Safeworld to solve it.
"The safety challenge we're talking about is a combination of, one, really advanced generative AI probabilistic evals—how do you underwrite the risk of a probabilistic system?" Zhao says. "The second part that's really hard is the trust part, and you need both to deploy a robot."
Seed Funding and Backers
Safeworld is emerging from stealth today with a seed round of more than $12 million, led by Shine Capital and a16z Speedrun, with additional investment from Box Group, Carnegie Mellon University Endowment, Innovation Endeavors, and SV Angel.
"The time to build an industry safety standard is now, while robots are being designed and deployed," a16z Speedrun partner Jonathan Lai told TechCrunch. "By the time you have robots in households colliding with kids and causing safety incidents, that's way too late."
Simulation as the Core Safety Engine
Safeworld's specialty is evaluating a robotic control system in simulations populated with realistic human models. It's akin to the challenge faced by companies like Tesla or Wayve, which must ensure their vehicles respond appropriately to a variety of surprising incidents on the road. But it will be more difficult for robots, Zhao argues, because they work in unstructured environments—and because each facility they operate in will have different safety standards.
"One of the most common areas is if there is a blind corner in this particular factory," Wong said. "What is the speed or what is the stopping distance that you need to make sure that this robot will not collide with a particular human? If a human is carrying boxes, for example, will the robot detect the human or not?"
To answer that question, Safeworld will build a digital version of that corner in a model like Genesis or MuJoCo, insert a simulation of the robot it is evaluating, driven by its real software, and then run thousands of scenarios where human models encounter the robot. That's harder than it seems, per Zhao, because people are unpredictable.
"Tripping and falling is also a good example of something that we do a lot of testing with in the simulation," Wong said. "Otherwise, you would have to go and trip and fall for the robot, which is a hard thing to be doing all the time."
Third-Party Validation and Industry Standards
There are clear similarities between the platform Safeworld is building and the tools being used internally by robot builders. The founders, however, believe that beyond their specific expertise, robot-makers will want a third party to validate their work—if only to share information about safety cases between competitors.
"A lot of people are underestimating how hard some of these edge cases are going to be to solve," Zhao said. "It is not the robot in the vacuum, in the demo, that we are worried about. It is the robot that is deployed at scale, with people who potentially never operated a robot before."
Early Partners and Real-World Deployment
Vishal Dugar, the CTO of Gritt Robotics, is developing the AI brain for robots that currently help workers install photovoltaic panels at industrial-scale solar farms, and aspires to take on more complex construction tasks. His company is partnering with Safeworld as they develop their safety simulations.
As of 2026, the robotics industry is at an inflection point: generative AI models are being embedded into physical systems at an accelerating pace, and regulators are beginning to take notice. Safeworld's bet is that safety and trust will become the bottleneck for mass deployment—and that a dedicated, third-party validation layer will be essential infrastructure for the entire field.
via TechCrunch AI