Three years ago, Wirestock was a marketplace where photographers and illustrators sold stock images. Today it's a training data supplier - and it just raised $23 million to keep building that business.
Wirestock made its pivot in 2023, shifting from creative marketplace to AI data provider. It now packages images, videos, design assets, and gaming and 3D content into structured datasets for AI labs. Training data is the raw material machine learning models learn from - a model can only generate or understand content types it's seen examples of during training. Labs building multimodal models (ones that work across images, video, and 3D rather than text alone) need large quantities of diverse visual data, and sourcing it in volume is genuinely hard.
The Tension in the Business Model
Wirestock built its original community around professional creators who licensed their work commercially. Those same creators now compete with AI image tools trained partly on datasets like Wirestock's. The company sits at the middle of that chain: aggregating human creative work, then selling it to systems that reduce demand for that same work.
How Wirestock handles contributor rights and compensation will matter as this tension becomes more public. Several AI image companies have already faced legal challenges over training data sourcing. Suppliers who can document clean rights chains for their datasets have a real advantage as the legal landscape tightens.
The $23M lets Wirestock expand at a moment when lab demand is high. Its coverage - photography, design, gaming content, and 3D material - gives it range that a pure stock-photo dataset wouldn't. For companies building products on top of AI image or video generation, training data quality determines what those models can do. Wirestock is betting that AI labs will keep needing better data faster than they can source it themselves.