The Google and Nvidia-backed AI video startup Runway is partnering with Getty Images to develop Runway Getty Images Model (RGM), which it is positioning as a new type of generative AI model capable of “providing a new way to bring ideas and stories to life through video” for enterprise customers using copyright compliant means. Targeting Hollywood studios, advertising, media and broadcast clients, RGM will “provide a baseline model upon which companies can build their own custom models for the generation of video content,” Runway explains.

Runway CEO and co-founder Cristóbal Valenzuela says he expects RGM to “unlock new commercial uses and new video products,” suggesting in a blog post that it will empower “a new level of creative control and customization” for creators.

“Enterprise customers will be able to fine-tune RGM using their own proprietary datasets,” powering new content workflows for video creation that in addition to storytelling can be “tailored to style and brand identities,” he explains.

VentureBeat says the move is notable on several fronts, most obviously by “setting its sights on penetrating some of the biggest visual media industries in the world and becoming an entrenched part of the workflow,” newly adapting to AI in those sectors.

Getty and its royalty-supported data library also lends legitimacy to Runway, VentureBeat adds, noting that the New York-based AI firm is currently the target of a copyright lawsuit. However, VentureBeat contends that Getty competitor Adobe Stock has had contentious discussions with contributors over use of their images “without express notification or consent” to train the company’s Firefly AI models.

“The landscape of generative text-to-video technology has rapidly changed in the last few weeks,” PetaPixel writes, citing Stability AI announcing Stable Video Diffusion and launch of a 1.0 generative video platform from Pika Labs, which just received $55 million in funding.

Observing that generative video content “is much more resource-intensive and challenging than creating still images,” PetaPixel predicts that “generative video will undoubtedly improve significantly in 2024,” much like single-frame generative image technology has in 2023. PetaPixel in March proclaimed Runway Gen2 “the first publicly available generative text-to-video generator.”

The Getty partnership follows Runway’s June release of Gen-2 and web-based software, with improvements “including a new motion brush that allows users to add select motion to different areas of a static image, new camera motions (that simulate a video camera’s POV allowing users to pan and tilt around a static image, adding cinematic motion to it), and an upgraded fidelity and realism of the resulting videos,” per VentureBeat.