Nvidia is expanding a safety system used in self-driving cars to humanoid and industrial robots as it steps up efforts in physical AI. The company has widened its Halos platform to cover autonomous mobile robots, factory humanoids and surgical robots, integrating computing modules, dedicated safety processors and software monitoring. It also offers simulation and testing tools. Nvidia says safety is becoming a bottleneck as robot capabilities improve, and the platform lets developers tailor protections to different environments.

Netherlands-based security startup Hadrian Security has raised 36 million euros in funding, a tech publication reported on Monday. The company provides AI agents that identify vulnerabilities attackers could actually exploit. Its platform includes Atlas, which monitors external attack surfaces and verifies exploitability, and Nova, which runs on-demand penetration tests. Hadrian plans to use the funds to expand overseas and grow engineering and research teams, targeting nine-figure revenue long term.