At TechCrunch Disrupt 2026, leaders from Shield AI, Waabi, and General Motors explored the critical challenges of developing AI systems where failure is not an option. The panel addressed how AI autonomy in aircraft, vehicles, and robots demands rigorous safety cultures, extensive testing, compliance with regulations, and building user trust before deployment. Their insights highlighted the unique stakes involved when AI transitions from controlled environments to real-world, high-risk applications.

Nathan Michael, CTO of Shield AI, shared his experience leading the Hivemind autonomy platform, which is deployed in highly sensitive defense settings. With a background in robotics and AI from Carnegie Mellon, Michael discussed technology designed to meet strict assurance requirements, including Shield AI’s recent selection by the U.S. Air Force for its Collaborative Combat Aircraft drone program. The company’s recent $1.5 billion Series G funding round, valuing it at $12.7 billion, underscores the significance of its defense-focused AI efforts.

Raquel Urtasun, founder and CEO of Waabi, brought a deep autonomous driving expertise honed over 25 years, including leadership at Uber ATG and the University of Toronto. Waabi’s approach centers on thorough validation and simulation testing, with their Waabi World platform preparing autonomous trucks and robotaxis for the road. In January 2026, Waabi raised $1 billion and partnered with Uber to deploy 25,000 robotaxis, emphasizing that full validation is essential before any driverless deployment.

General Motors’ Director of Robotics Strategy, Mikell Taylor, contributed insights from her work building practical robotics designed to collaborate safely with humans. Taylor emphasized the importance of user experience and workforce trust as autonomous systems move from demonstration phases into everyday industrial and commercial use. This conversation at Disrupt underscored that for AI systems in critical domains, “almost ready” is not sufficient; readiness means achieving reliability and safety that users and operators can fully depend on.