First there were XR studios, then it was “metaverse” studios, or even “virtual” studios. Whatever the tech buzzword, there have always been a new wave of upstart companies, dedicated to being the leaders in creating content for whatever the new medium was at that moment.
Today the buzzword — and the one that actually seems like it’s more than just a fad — is generative artificial intelligence. The wave of people touting themselves as the new studio tastemakers with Gen-AI has been impossible to keep track of.
Well, someone finally did.
A consultancy group and product innovation lab called FBRC.ai on Friday released a report that aims to quite literally map out all the AI content creators in the landscape as of March 2025. It can’t profess to be completely comprehensive (another AI studio just launched as you finished reading this sentence), but the group’s inaugural study received survey responses from 98 different companies that self-identify as AI studios, even if it’s just “the tip of the iceberg,” as the report says, of all the studios that could be out there.
Of those 98 studios, 65 have launched since 2022, and over 30 have launched in just 2024 and the first two months of 2025 alone. The report found that many of these companies have just 5 or fewer employees and are largely independent endeavors. It speaks to how rapidly the industry around Gen-AI is growing and how many want to be the first to truly break out in the mainstream.
In addition to a broad overview of the space and a world map of where each of these companies are based, the full report includes profiles with nine different studios who are among the current leaders in the space: Promise, Asteria, Shy Kids, Invisible Universe, Playbook, AI Hub, Pigeon Shrine, Secret Level, and Mod Tech Labs.
As the report shows, not all of these companies are created equal, and they exist on a spectrum in terms of how they utilize AI. Some are work-for-hire agencies that almost solely bring in clients, while others co-produce, and some are working to develop entirely their own content they can create and distribute.
What’s more, some companies are considered “AI Native” studios, in that they use “closed source” models like OpenAI’s Sora, Runway, or Luma’s Dream Machine and will create text, images, and video entirely via AI prompts going direct from script to screen. Others will use an open source model like ComfyUI that gives creators a little more control over the creation tools. And others still are merely using AI to augment traditional production pipelines, making either live-action/AI hybrids or just using AI to be more efficient.
“Part of the way that the conversation around AI in media and entertainment has progressed positively is that the voices speaking in the space now have experience with actually telling stories,” the report by Rachel Joy Victor reads. “While all have experimented with technology throughout their careers and thus were drawn to AI as the next frontier, they all had years of experience with storytelling. Nearly every founder attests that they still believe the most substantial moat is not technical but rather about who can do the work of crafting a good story.”
Making these distinctions in the AI space is important for people who have a narrow or pigeon-holed idea of what generative AI production is or what it looks like. But the report acknowledges that, over time and as the models improve, many of these lines will start to blur, with studios experimenting with hybrid projects and fully-blown AI movies.
With the exception of Asteria, which just launched their own “clean model” called Marey, most of the companies profiled are not building their own AI models and will leave that work to the tech giants like OpenAI and Runway. But they are differentiating themselves by developing their own datasets on which these models are trained (think Lionsgate’s recent partnership with Runway), as well as developing their own unique workflows or interfaces about how they want to use the tools. According to Playbook’s Skylar Thomas via the report, that’s because savvy viewers are starting to notice that individual AI models each have their own look to them, and using them via your own dataset is a way to stand apart.
So are all these companies working to take everyone’s jobs? Not necessarily, the report concludes. While some are dabbling in shorts, feature films, and TV shows, others are invested in VR, gaming, ads, and music videos, among other new formats of content. And while it may be cheaper and faster to produce content, not all of these small companies have the infrastructure necessary to support elaborate models and create sustained business models.
Check out the full report from FBRC.ai here.
