TechCrunch: Runway moves from AI video generation to world models

Runway, founded in 2018 by three NYU Tisch School of the Arts graduates, has been a fixture in AI-assisted creative workflows through its Gen-4 video generation models. The company’s tools are used by filmmakers, advertising agencies, and creative studios for everything from motion graphics to AI-assisted film production. A May 15, 2026 TechCrunch profile describes where the company is heading next.

The shift is toward world models — AI systems designed to simulate environments and predict physical behavior, rather than generate individual media outputs. Where language models predict text and image models predict pixels, world models predict how objects move, how environments change, and how agents interact with space. Runway launched its first world model in December 2025, with another planned for 2026.

The company is also entering robotics through a dedicated unit with real-world deployments already in place, and expanding into drug discovery and climate modeling. These applications are well outside the design studio, but they reflect where the underlying technology is being directed.

For creative professionals who currently use Runway for motion graphics, video concepting, or visual prototyping, the direction suggests that future tools built on world model infrastructure could understand physics and spatial relationships rather than only generate plausible-looking footage. That would change what AI can contribute to design work involving motion, interaction, or three-dimensional space.

Runway’s stated ambition is to build foundational AI infrastructure rather than remain a product company serving a single industry. The timeline for when that infrastructure translates into tools that reach designers in production workflows is not clear from the article, but the direction of investment is.