Origin Lab raises $8M to assist video game corporations sell data to world-model builders

As AI begins to interact with the physical world, latest varieties of labs are working to construct world models that may very well be used to operate physical robotics or model objects in physical space. Unlike large language models, there isn’t a simple source of information for those models, which has left many labs scrambling to assemble the crucial training sets.

Now, one startup is emerging with an unlikely data source: the video game industry.

That’s the premise of Origin Lab, which just announced an $8 million seed funding round led by Lightspeed Ventures. SV Angel, Eniac, Seven Stars, and FPV also participated, with angel funding from Twitch co-Founder Kevin Lin and Cruise founder Kyle Vogt.

“The AI systems which are being built now need to grasp how the physical world works and the way things move,” co-CEO and co-founder Anne-Margot Rodde told TechCrunch. “That data essentially lives in video games.”

In easy terms, Origin Lab will function a marketplace where world-model-focused labs similar to Yann LeCun’s AMI Labs or Fei-Fei Li’s World Labs should buy high-quality licensed data. On the opposite side of the trade, video game corporations can squeeze additional revenue out of the digital assets they’ve already created. In the center, Origin Lab will convert the video game assets right into a form that works as training data — something that may very well be so simple as a rendering run or as complex as automating hours of walkthrough footage.

“It became clear that the video game industry was sitting on some incredibly beneficial data, but there was no possible way or infrastructure to principally connect AI labs and the video game industry,” says Rodde. “So essentially, we built that bridge.”

Labs have long been serious about video game footage as an information source, but licensing and data quality issues have often gotten in the way in which. In December 2024, OpenAI caused a minor scandal when the primary version of its Sora video-generation model appeared to regurgitate footage of popular video games and streamers — presumably since it had been trained on Twitch streams. Amazon has been open about its interest in using Twitch footage to coach models.

Origin’s success in fundraising is an indication of a growing market — not only for training data, but for startups that may function essential suppliers to major AI labs. Faraz Fatemi, a partner at Lightspeed who led the Origin investment, says the success of corporations like Scale.AI has made the chance not possible to disregard.

“We’ve seen how sharp the revenue scaling will be for data vendors which are serving the key labs,” Fatemi told TechCrunch. “These are very well-capitalized businesses, and the bottleneck for all of them is data.”