Origin Lab Raises $8M to Turn Video Games Into AI Training Data

As artificial intelligence companies race to build advanced world models capable of understanding the physical world, one startup believes the answer may already exist inside video games.

Origin Lab has raised $8 million in a Seed funding round led by Lightspeed Venture Partners to help AI labs access high-quality training data sourced from the gaming industry.

The funding round also included participation from SV Angel, Eniac Ventures, Seven Seven Six, and First Person Ventures. Angel investors included Kevin Lin and Kyle Vogt.

The startup is betting that video game environments could become one of the most valuable resources for training next-generation AI systems designed to interact with the real world.

“The AI systems being built today need to understand how objects move and how the physical world behaves,” said Anne-Margot Rodde in an interview with TechCrunch. “That kind of data already exists inside video games.”

Rodde founded the company alongside Antoine Gargot and Colin Carrier.

Origin Lab plans to act as a marketplace connecting video game studios with AI labs building world models.

These world models are advanced AI systems capable of simulating environments, understanding physical interactions, and potentially powering robotics or future autonomous systems.

The company says labs such as World Labs, founded by Fei-Fei Li, and AMI, associated with Yann LeCun, could benefit from licensed gaming data for training purposes.

On the other side, video game developers gain a new way to monetize digital assets they have already spent years building.

Origin Lab’s role is to convert raw gaming assets into usable AI training data. That process could involve anything from rendering 3D environments to generating detailed gameplay walkthrough footage for machine learning systems.

Interest in gaming data for AI training has been growing rapidly in recent years. However, licensing problems and data-quality concerns have often prevented broader adoption.

The issue gained attention in late 2024 when OpenAI faced criticism after early versions of its Sora model appeared to reproduce scenes from popular video games and livestream content, leading to speculation that Twitch footage may have been used during training.

Amazon has also publicly explored the idea of using content from Twitch to improve AI systems.

Investors believe startups focused on AI infrastructure and training data could become critical suppliers for the rapidly expanding AI industry.

Faraz Fatemi, who led the investment, said companies providing specialized training data are becoming increasingly valuable as demand from AI labs continues rising.

“We’ve seen how quickly data infrastructure businesses can scale,” Fatemi told TechCrunch. “For major AI labs, data remains one of the biggest bottlenecks.”

With world-model development becoming a major focus across the AI sector, Origin Lab is positioning itself at the intersection of gaming, artificial intelligence, and next-generation robotics infrastructure.

Also read : Vapi Hits $500M Valuation After Winning Amazon Ring Deal